# Frequently asked questions about GenRank
Source: https://docs.genrank.io/account/faq
Answers to common questions about how GenRank tracks AI visibility, what the data means, how it differs from SEO, and what each plan includes.
GenRank measures how your brand appears in ChatGPT responses — what it tracks, how it works, and what you can do with the data. Here are answers to the questions we hear most often.
A ChatGPT rank tracker measures how visible your brand is inside ChatGPT answers for specific prompts. Instead of tracking positions in a search engine results page, it analyzes how ChatGPT mentions, positions, and references your brand in actual responses.
GenRank runs your selected prompts in ChatGPT on a recurring basis, captures the full outputs, and identifies where and how your brand appears — alongside competitor mentions, source citations, and the attributes the model associates with your name. The result is a daily picture of your brand's presence in the AI discovery channel.
All data comes from real ChatGPT responses. GenRank executes your tracked prompts directly in ChatGPT, captures the complete outputs, and structures the data from those actual responses. There are no estimates, simulated outputs, or extrapolations.
Each response is stored so you can review the exact text where your brand was mentioned or cited. You see the full ChatGPT output, highlighted mentions, the citations table, and the associated brand perception signals — all from verified, captured responses.
GenRank fetches responses for every tracked prompt daily. Data is refreshed every 24 hours across all plans, including the Free plan. This lets you monitor trends, detect shifts in visibility, and validate the impact of content or PR changes over time.
Yes. Competitor tracking is included on Essential, Pro, and Scale plans. GenRank shows how competitor brands perform across the same prompts you track — which ones dominate, which prompts they own, and where your brand is absent while theirs appears.
Competitor limits by plan:
* **Essential** — 3 tracked competitors
* **Pro** — 10 tracked competitors
* **Scale** — custom
The Free plan does not include competitor tracking.
Keywords are short, fragmented search terms — for example, "project management software." Prompts are the natural, contextual questions or instructions that people actually type into ChatGPT — for example, "what's the best project management software for a remote team of 10?"
AI systems respond to prompts, not keyword strings. Tracking keywords against ChatGPT does not reflect real usage. Prompts mirror actual conversational behavior, which is why GenRank is built around prompts as the core tracking unit. Converting existing keyword research into realistic prompts is one of the first steps when setting up a project.
AI visibility is context-dependent. Your brand may appear prominently in one type of query and be completely absent in another. The prompts you track determine every metric in GenRank — mention rate, share of voice, competitor gaps, perception data, and content requirements are all scoped to the prompts you define.
Poor prompt selection means your data reflects questions no one is actually asking. Strong prompt selection, built from Page Scan outputs, Google Search Console queries, or deliberate manual curation, ensures you track the conversations that actually influence discovery and buying decisions in your category.
SEO tools track your position in Google's ranked list of links. GenRank tracks your presence inside AI-generated answers, where there is no ranked list — only the brands ChatGPT chooses to mention.
The underlying signals are also different. SEO optimization targets keywords, backlinks, and crawlability. GenRank's optimization focuses on retrieval alignment, entity resolution, and the structural clarity of your content inside AI systems. These are complementary practices — GEO does not replace SEO, it adds a measurement layer for an increasingly important discovery channel.
Yes. Prompts can be added in bulk and managed centrally in the Prompt Manager. All plans above Free include meaningful prompt slots, and the Scale plan offers unlimited custom prompts with usage-based billing.
The prompt landscape also evolves — as products, competitors, and user behavior change, your tracking should keep up. Prompt Research in GenRank is designed as an ongoing process, not a one-time setup. Page Scan and Google Search Console integration give you a repeatable way to surface new prompts as your category shifts.
Yes. For every tracked prompt, you can review the full ChatGPT response where your brand — or a competitor — was mentioned. Each stored response includes the complete output text, highlighted brand mentions, the citations table, and context on brand perception signals extracted from that response.
You are not seeing summaries or reconstructions. You are seeing the actual captured output from a real ChatGPT session.
Yes. By aligning your content with the patterns that appear in pages AI already cites, strengthening entity clarity so your brand is resolved consistently, and addressing the external sentiment signals that shape AI's perception, you directly influence which information gets selected and how your brand is framed in responses.
GenRank's Optimization module translates AI responses into concrete, page-checkable requirements. Acting on those requirements — improving content structure, fixing entity fragmentation, addressing negative sentiment sources — changes what AI retrieves and how it describes your brand over time.
Optimization is a continuous process. AI systems evolve, competitors update their content, prompt landscapes shift, and new external sources emerge. A one-time audit gives you a point-in-time snapshot, but it goes stale quickly.
GenRank monitors content alignment, entity fragmentation, and sentiment signals on an ongoing basis so you can detect when changes occur and respond before they compound into larger visibility losses.
Generative engine optimization is the practice of improving how your brand appears in AI-generated answers, specifically in tools like ChatGPT. Where SEO targets search engine rankings, GEO targets the selection and framing of brands inside AI responses.
GEO involves three interconnected practices: identifying the prompts that define your category's AI conversations (Prompt Research), measuring your brand's performance across those prompts (Response Tracking), and improving the structural, semantic, and reputational signals that influence AI selection (Optimization). GenRank is built to support all three.
GenRank currently tracks visibility in ChatGPT, which accounts for the majority of AI tool usage worldwide. Support for additional large language models and AI search tools is planned.
You can follow product updates on the [GenRank blog](https://genrank.io/blog) or [contact the team](https://genrank.io/contact-us/) to ask about upcoming integrations.
# Plans and pricing for GenRank
Source: https://docs.genrank.io/account/pricing
Compare GenRank's four plans — Free, Essential, Pro, and Scale — to find the right fit for your brand tracking needs and budget.
GenRank offers four plans priced in EUR, from a free tier for getting started to a fully custom Scale plan for agencies and enterprise teams. Every paid plan includes daily data refresh so your visibility data is never more than 24 hours old.
## Plan comparison
| | Essential | Pro | Scale |
| ---------------- | --------- | ------- | --------- |
| **Price** | \$69/mo | €199/mo | Custom |
| **Prompt slots** | 25 | 150 | Unlimited |
| **Projects** | 1 | 3 | Custom |
| **Features** | Basic | All | All |
## What's included in each plan
Get your first look at AI visibility without a credit card. Track up to 10 prompts across 1 project and see where your brand appears in ChatGPT responses.
**Features:** Brand Visibility
**Limits:** 10 prompt slots · 1 project · Limited data retention
For business owners who need competitive context alongside their own brand data.
**Features:** Brand Visibility, Competitor Insights, Sources, Brand Perception
**Limits:** 25 prompt slots · 1 project · 3 tracked competitors
For brands and specialists running a defined GEO strategy. Includes the full Optimization module and support for multiple projects.
**Features:** All Essential features + Content Optimization, Entity Clarity
**Limits:** 150 prompt slots · 3 projects · 10 tracked competitors
For agencies and enterprise teams that need custom prompt volumes, unlimited projects, API access, and dedicated support.
**Features:** All Pro features + API Access, Priority Support
**Limits:** Custom prompt slots · Custom projects · Custom competitors
## Feature details
### Brand Visibility
Available on all plans. Tracks how often your brand is mentioned and cited across your tracked prompts, updated daily.
### Competitor Insights
Available on Essential and above. Shows how competitor brands perform across the same prompts — which ones dominate, which are fading, and where you have visibility gaps.
### Sources
Available on Essential and above. Identifies the external domains that are cited in AI responses about your category, and shows how your own domain performs within that landscape.
### Brand Perception
Available on Essential and above. Extracts the recurring qualities and themes that ChatGPT associates with your brand across responses, so you understand how AI positions you.
### Content Optimization
Available on Pro and above. Analyzes the content patterns behind pages that get cited, translates those patterns into concrete page-level requirements, and shows how your content aligns at the vector-similarity level.
### Entity Clarity
Available on Pro and above. Measures how consistently ChatGPT resolves your brand as a single, unambiguous entity. Detects naming variations and fragmentation before they weaken recognition.
### API Access
Available on Scale only. Query your GenRank data programmatically and integrate visibility metrics into your own dashboards or workflows.
Scale plan pricing is custom and based on your prompt volume, number of projects, and team size. [Contact us](https://genrank.io/contact-us/) to get a quote or book a demo.
## All paid plans include daily refresh
GenRank runs your tracked prompts in ChatGPT every day and captures the full responses. You always have up-to-date data without any manual re-runs. The Free plan also refreshes daily, but data retention is limited.
Not sure which plan to start with? Begin on Free to verify that GenRank tracks your brand correctly, then upgrade to Essential or Pro once you're ready to add competitor tracking or optimization tools.
## Get started
[Start for free](https://app.genrank.io/onboarding) — no credit card required. You can upgrade at any time from **Account → Billing** inside the dashboard.
# Projects: brand tracking configurations
Source: https://docs.genrank.io/account/projects
A GenRank project defines what you track — your brand, domain, geo target, competitors, and prompt set. Learn how to set one up and when to use multiple.
A project in GenRank is a self-contained tracking configuration. It tells GenRank which brand to monitor, which website domain to associate with it, which geographic market to focus on, which competitors to benchmark against, and which prompts to run in ChatGPT. All of your visibility data — mentions, perception, competitors, sources — is scoped to a project.
## What a project contains
Each project has five core settings:
| Setting | What it defines |
| ------------------- | ---------------------------------------------------------- |
| **Brand name** | The name GenRank listens for across ChatGPT responses |
| **Website domain** | Your primary domain, used for source and citation tracking |
| **Geo targeting** | The market context for prompt execution (e.g., US, EU, UK) |
| **Competitor list** | The brands you want to benchmark against |
| **Prompt set** | The prompts GenRank runs in ChatGPT on your behalf |
Use geo targeting to measure how your brand performs in specific markets. A brand that appears in UK responses may be absent in US ones — or vice versa. Running separate projects per region gives you a clear, unambiguous view of each market rather than mixing data across them.
## How to create a project
You set up your first project during onboarding. Additional projects (on Pro and Scale plans) can be created directly from your dashboard.
If you're a new user, the onboarding flow walks you through project creation automatically. If you already have an account, open the dashboard and select **New Project** from the project switcher.
Type the brand name exactly as you want GenRank to detect it in ChatGPT responses. Add your website domain so source and citation data is attributed correctly.
Choose the geographic market you want to track. This scopes how prompts are executed and ensures visibility data reflects the right regional context. You can change this setting later.
Enter the names of the competitor brands you want to benchmark against. The number of tracked competitors depends on your plan: 3 on Essential, 10 on Pro, and custom on Scale.
Add the prompts GenRank will run in ChatGPT each day. You can create prompts manually in the Prompt Manager, generate them from a URL using Page Scan, or import them from your Google Search Console data.
## Project limits per plan
| Plan | Projects | Tracked Prompts |
| --------- | -------- | --------------- |
| Essential | 1 | 25 |
| Pro | 3 | 150 |
| Scale | Custom | Custom |
Free and Essential plans include a single project. If you need to track more than one brand or market simultaneously, upgrade to Pro or Scale.
## When to use multiple projects
Multiple projects are most useful when you need visibility data to stay clearly separated. Common reasons to run more than one project:
**Multiple brands.** If you manage more than one brand — whether as an in-house team or an agency — each brand should have its own project. Mixing brands in a single project makes it impossible to isolate their individual performance.
**Multiple geographic markets.** The same brand can have very different ChatGPT visibility in the US, EU, and UK. A separate project per region lets you track each market independently, compare results side by side, and prioritize where to invest in content or GEO improvements.
**Different product lines.** If your company operates in distinctly different categories — each with its own competitor set and relevant prompts — separate projects help you keep those category signals from bleeding into each other.
**Agency use cases.** Agencies managing AI visibility for multiple clients use projects to keep each client's data isolated, making reporting clean and client-specific.
If you're an agency or managing a multi-brand portfolio, the Scale plan gives you a custom number of projects with no fixed ceiling. [Contact us](https://genrank.io/contact-us/) to discuss your setup.
# API Authentication: Securing Your GenRank Access
Source: https://docs.genrank.io/api/authentication
Learn how GenRank API credentials are provisioned during Scale plan onboarding and how to authenticate every API request with a Bearer token.
GenRank API access is provisioned by the GenRank team as part of your Scale plan onboarding. You do not generate credentials yourself through a self-serve dashboard — instead, your API key and endpoint details are delivered to you once your integration has been configured and your account is ready.
This approach exists because every GenRank API setup is tailored: your endpoint, your tracked brand configuration, your custom output fields. Provisioning credentials manually ensures your integration is verified and correctly scoped before you start sending requests.
Your specific API key, base URL, and endpoint details are provided by GenRank during onboarding. If you have not yet started the process, [contact GenRank](https://genrank.io/contact-us/) to get set up on the Scale plan.
## How authentication works
The GenRank API uses **Bearer token authentication**. Every request must include your API key in the `Authorization` header as a Bearer token. Requests without a valid token are rejected with a `401 Unauthorized` response.
### Required headers
Your API key passed as a Bearer token. Format: `Bearer YOUR_API_KEY`.
Must be `application/json` for all requests.
## Example authenticated request
The following example shows how to structure an authenticated request using cURL. Replace `YOUR_API_KEY` with the key provided during your onboarding.
```bash theme={null}
curl -X GET "https://api.genrank.io/v1/mentions" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json"
```
The same pattern applies in any HTTP client or SDK. Here are equivalent examples in Python and JavaScript:
```python python theme={null}
import requests
headers = {
"Authorization": "Bearer YOUR_API_KEY",
"Content-Type": "application/json"
}
response = requests.get("https://api.genrank.io/v1/mentions", headers=headers)
data = response.json()
```
```javascript javascript theme={null}
const response = await fetch("https://api.genrank.io/v1/mentions", {
method: "GET",
headers: {
"Authorization": "Bearer YOUR_API_KEY",
"Content-Type": "application/json"
}
});
const data = await response.json();
```
## Credential delivery
When your GenRank Scale plan is set up and your API integration is configured, you receive:
* **Your API key** — a secret token that authenticates your requests
* **Your base URL** — the endpoint specific to your account configuration
* **Integration notes** — any custom fields, parameters, or request patterns specific to your setup
Store these details securely. If you lose access to your key or need to rotate it, contact GenRank support.
## Keeping your API key secure
Never commit your API key to version control. If your key is exposed in a public repository or shared accidentally, contact GenRank immediately to have it rotated.
Follow these practices to keep your credentials safe:
* Store your API key in environment variables, not hardcoded in source files
* Use a secrets manager (such as AWS Secrets Manager, HashiCorp Vault, or GitHub Secrets for CI/CD) to inject the key at runtime
* Restrict access to the environment variable or secret to only the services that need it
* Never log the full `Authorization` header or print your key to console output
* Rotate your key periodically, or immediately if you suspect it has been compromised
```bash theme={null}
# Store your key in an environment variable
export GENRANK_API_KEY="your_api_key_here"
# Reference it in requests
curl -X GET "https://api.genrank.io/v1/mentions" \
-H "Authorization: Bearer $GENRANK_API_KEY" \
-H "Content-Type: application/json"
```
## Authentication errors
If your request is not properly authenticated, the API returns a `401 Unauthorized` response. Common causes include:
* Missing `Authorization` header
* Malformed token (e.g., using `Token` instead of `Bearer`)
* Expired or revoked API key
* Using a key from a different account or environment
If you are consistently receiving authentication errors after confirming your header format is correct, contact GenRank support to verify that your key is active and properly scoped to your account.
# Brand Data: Configuration Fields in the API
Source: https://docs.genrank.io/api/brand-data
Understand the brand configuration fields in every GenRank API response — brand name, domain, geo targeting, and the last_checked timestamp.
Every GenRank API response is anchored to a brand configuration: the brand name, domain, and geographic target that you defined when setting up your GenRank integration. These fields appear at the top level of every response object and tell you exactly which brand, which market, and which prompt run the data belongs to.
Understanding these fields is important when you are working with multiple brands, multiple geographies, or when you need to correctly label and route incoming data to the right part of your downstream system.
## Brand configuration fields
The following fields appear in every top-level API response object.
The name of the tracked brand as configured in your GenRank account. This is the string GenRank uses when detecting brand mentions in ChatGPT responses.
Example: `"GenRank"`
The primary domain associated with the tracked brand. GenRank uses this field to identify citation mentions — instances where ChatGPT links or refers to your website specifically.
Example: `"genrank.io"`
The geographic market for which the prompt was executed. GenRank runs prompts in a way that simulates a user from the specified region, so results reflect what ChatGPT returns for users in that market.
Common values include `"US"`, `"UK"`, `"EU"`, `"AU"`, and other ISO country or region codes. Your available geo targets depend on your account configuration.
Example: `"US"`
The exact prompt text that was submitted to ChatGPT for this data point. Each response object corresponds to a single prompt run, so this field identifies which question generated the mention and visibility data in the response.
Example: `"best chatgpt brand tracker"`
An ISO 8601 timestamp indicating when GenRank last executed this prompt and captured the response. This field tells you the freshness of the data — GenRank runs prompts daily, so this value should typically reflect the most recent daily run.
Example: `"2025-09-26T10:00:00Z"`
## Sample response (brand fields)
The following shows a GenRank API response with brand configuration fields populated. The `mentions` array is truncated here — see [Mentions](/api/mentions) for full field documentation.
```json theme={null}
{
"brand": "GenRank",
"domain": "genrank.io",
"geo": "US",
"query": "best chatgpt brand tracker",
"mentions": [...],
"last_checked": "2025-09-26T10:00:00Z"
}
```
## What geo targeting means
The `geo` field is not just a label — it determines how GenRank executes the prompt. When a geo target is set to `"US"`, prompts are run in a context that simulates a US-based user asking ChatGPT. This matters because ChatGPT's responses can vary by region: different brands are recommended, different sources are cited, and local market context influences the output.
If you are tracking a brand across multiple markets, you will receive separate response objects per geo — one for each market you have configured. Use the `geo` field to route or segment that data correctly in your downstream systems.
If you are building a reporting dashboard that covers multiple geographies, always group and filter by the `geo` field rather than assuming all responses represent the same market.
## Tracking freshness with `last_checked`
The `last_checked` timestamp is your signal for data freshness. GenRank runs all tracked prompts daily, so under normal operation, this value will be no more than 24–48 hours old.
If you are storing GenRank data in your own database and want to avoid processing stale duplicates, compare the incoming `last_checked` value against the most recent stored value for the same `brand` + `geo` + `query` combination. Only store the record if the timestamp is newer.
```python theme={null}
from datetime import datetime
def is_fresh(existing_checked: str, incoming_checked: str) -> bool:
existing = datetime.fromisoformat(existing_checked.replace("Z", "+00:00"))
incoming = datetime.fromisoformat(incoming_checked.replace("Z", "+00:00"))
return incoming > existing
```
## How brand settings map to what is tracked
Your brand configuration is set once during onboarding and serves as the foundation for every data point GenRank collects. If you need to update your brand name, domain, or geo targets — for example, because you are expanding into a new market or rebranding — contact GenRank to update your configuration. Changes take effect from the next daily data run.
GenRank tailors the integration to your needs. If you want to track multiple brands, multiple domains, or additional geographic markets beyond your current setup, reach out to discuss expanding your configuration.
# Competitor Data: Share of Voice via the API
Source: https://docs.genrank.io/api/competitors
How competitor co-occurrence data surfaces in GenRank API responses and how to use it for share-of-voice calculations, dashboards, and spike alerts.
Every GenRank API mention object includes a `competitors` field listing the other brands that appeared in the same ChatGPT response as your brand. This co-occurrence data is the foundation for competitive analysis through the API — it tells you which brands share the response space with you, and by extension, which ones are competing for AI visibility on the same prompts you are targeting.
Combined with your own visibility scores, competitor co-occurrence data lets you calculate share of voice, identify competitive threats early, and understand how your brand stacks up in AI-generated responses across your full prompt set.
## The `competitors` field
The `competitors` field is an array of strings inside each mention object. Each string is the name of another brand that appeared in the ChatGPT response that generated this mention.
An array of brand name strings representing other brands that co-occurred in the same ChatGPT response as your brand mention. If no other brands appeared in the response, this array will be empty.
The brands listed here are detected automatically by GenRank from the response text. They are not limited to brands you have configured as tracked competitors — any brand that appears in the response is captured.
Example: `["SEranking", "ChatGPT Tracker"]`
## Sample response with competitor data
```json theme={null}
{
"brand": "GenRank",
"domain": "genrank.io",
"geo": "US",
"query": "best chatgpt brand tracker",
"mentions": [
{
"position": 1,
"text_snippet": "GenRank is one of the top tools to track brand visibility in ChatGPT.",
"annotations": [
{
"entity": "GenRank",
"position_start": 0,
"position_end": 7,
"type": "brand"
}
],
"visibility_percent": 72.5,
"perceived_brand_qualities": [
"accurate",
"easy to use",
"SEO-focused"
],
"competitors": ["SEranking", "ChatGPT Tracker"]
}
],
"last_checked": "2025-09-26T10:00:00Z"
}
```
In this example, `"SEranking"` and `"ChatGPT Tracker"` both appeared in the same ChatGPT response as GenRank for the prompt `"best chatgpt brand tracker"`. Your brand appeared in position 1 with a visibility score of 72.5 — but the competitors field shows you are not the only brand being recommended.
## Use cases
### Building competitive dashboards
The `competitors` array gives you the raw data to build side-by-side competitive views. By aggregating competitor co-occurrence across all responses over time, you can calculate which competitors appear most frequently alongside your brand and which prompts generate the most competitive crowding.
```python theme={null}
from collections import Counter
def get_competitor_frequency(responses: list) -> dict:
competitor_counts = Counter()
for response in responses:
for mention in response.get("mentions", []):
for competitor in mention.get("competitors", []):
competitor_counts[competitor] += 1
return dict(competitor_counts.most_common())
# Example output:
# {"SEranking": 34, "ChatGPT Tracker": 19, "BrandWatch AI": 12}
```
Build a separate time-series view of competitor frequency per prompt. A competitor that barely appeared six months ago but now shows up consistently is gaining AI visibility momentum — early signal that you may need to respond with content or authority-building efforts.
### Calculating share of voice
Share of voice measures your brand's proportion of total brand mentions across all responses. The `competitors` array — combined with your own mention count — gives you everything you need to calculate this.
```javascript theme={null}
function calculateShareOfVoice(responses) {
const brandCounts = {};
responses.forEach(response => {
const brand = response.brand;
response.mentions.forEach(mention => {
// Count your brand
brandCounts[brand] = (brandCounts[brand] || 0) + 1;
// Count each competitor
mention.competitors.forEach(competitor => {
brandCounts[competitor] = (brandCounts[competitor] || 0) + 1;
});
});
});
const total = Object.values(brandCounts).reduce((sum, n) => sum + n, 0);
return Object.fromEntries(
Object.entries(brandCounts).map(([name, count]) => [
name,
((count / total) * 100).toFixed(1) + "%"
])
);
}
```
### Alerting when a competitor gains mentions
Monitoring competitor frequency over time allows you to set thresholds and trigger alerts when a specific competitor's co-occurrence rate rises above a baseline. This is useful for teams who want automated notifications when the competitive landscape shifts.
```python theme={null}
def check_competitor_spike(
current_counts: dict,
baseline_counts: dict,
threshold_percent: float = 20.0
) -> list:
alerts = []
for competitor, current in current_counts.items():
baseline = baseline_counts.get(competitor, 0)
if baseline > 0:
change = ((current - baseline) / baseline) * 100
if change >= threshold_percent:
alerts.append({
"competitor": competitor,
"change_percent": round(change, 1),
"baseline": baseline,
"current": current
})
return alerts
```
### Identifying prompts where competitors appear but you don't
Some of your tracked prompts may generate responses that include competitor brands but not yours. These are prompt gaps — high-value opportunities where competitors have AI visibility and you do not.
```python theme={null}
def find_competitive_gaps(responses: list, your_brand: str) -> list:
gaps = []
for response in responses:
has_your_brand = len(response.get("mentions", [])) > 0
if not has_your_brand:
# Collect any competitors mentioned in this response
all_competitors = set()
for mention in response.get("mentions", []):
all_competitors.update(mention.get("competitors", []))
# Even without your brand, check if competitors appeared elsewhere
# For gaps analysis, simply filter for empty mentions with competitor context
gaps.append({
"query": response["query"],
"geo": response["geo"],
"last_checked": response["last_checked"]
})
return gaps
```
## How competitor data complements brand visibility
Brand visibility data tells you how your own brand performs. Competitor co-occurrence data tells you who else is in the room. Together, they give you a more complete picture of how your category is being represented in AI-generated responses.
A high visibility score on a prompt where no competitors appear is a strong signal — you own that response space. A lower score on a prompt crowded with five competitors means you are sharing attention. A prompt where competitors appear but you do not is a gap to close.
Use visibility scores and competitor co-occurrence together, not in isolation, when making decisions about which prompts to prioritize for optimization.
The `competitors` field captures all brands that appeared in the response — not just the competitors you have explicitly configured in your GenRank account. This means you may discover emerging competitors in your category through the API before you are aware of them from other sources.
# Mentions: Brand Visibility Data in the API
Source: https://docs.genrank.io/api/mentions
Reference for the mentions array in GenRank API responses — position, text snippets, entity annotations, visibility scores, and perceived qualities.
The `mentions` array is the core of every GenRank API response. Each element in the array represents a single detected occurrence of your brand in a ChatGPT response — capturing where it appeared, what was said, how visible it was, what qualities were attributed to it, and which competitors appeared in the same response.
A single prompt execution can produce multiple mention objects if your brand is referenced more than once in the ChatGPT response. In most cases there will be one or two mention objects per response, but complex or longer responses may contain more.
## Mention object fields
An array of mention objects, each representing a detected brand mention within the ChatGPT response for the tracked prompt. An empty array indicates the brand was not mentioned in the response.
The position of this mention within the response — where in the sequence of brand references it falls. A value of `1` indicates this is the first mention of the brand in the response. Lower position values generally correlate with stronger prominence.
Example: `1`
An excerpt from the ChatGPT response containing the brand mention. This is the actual text as returned by ChatGPT — not paraphrased or summarized. Use this field to understand the context and framing of the mention.
Example: `"GenRank is one of the top tools to track brand visibility in ChatGPT."`
An array of entity annotation objects identifying specific named entities within the `text_snippet`. Each annotation marks the exact character position of the entity and classifies its type.
The name of the detected entity exactly as it appears in the text.
Example: `"GenRank"`
The zero-indexed character position in `text_snippet` where the entity begins.
Example: `0`
The zero-indexed character position in `text_snippet` where the entity ends (exclusive). You can use `text_snippet[position_start:position_end]` to extract the entity string directly.
Example: `7`
The entity type classification. For brand tracking, this value is typically `"brand"`. Additional types may be present depending on your configured outputs.
Example: `"brand"`
A score between `0` and `100` representing how visible the brand is within this response. The score accounts for factors including mention position, response length, and prominence of the mention relative to other brands. A higher score indicates stronger, more prominent visibility.
Example: `72.5`
An array of strings representing the qualities or attributes ChatGPT associated with your brand in this response. These are derived from the language used around your brand mention — adjectives, descriptors, and implied characteristics. Use this field to understand how AI currently frames your brand.
Example: `["accurate", "easy to use", "SEO-focused"]`
An array of strings listing other brand names that appeared in the same ChatGPT response as your brand. This field tells you which competitors co-occurred in the same answer — useful for competitive benchmarking and share-of-voice analysis. See [Competitors](/api/competitors) for more detail on how to use this data.
Example: `["SEranking", "ChatGPT Tracker"]`
## Sample mention object
The following shows a complete mention object as it appears in a GenRank API response.
```json theme={null}
{
"position": 1,
"text_snippet": "GenRank is one of the top tools to track brand visibility in ChatGPT.",
"annotations": [
{
"entity": "GenRank",
"position_start": 0,
"position_end": 7,
"type": "brand"
}
],
"visibility_percent": 72.5,
"perceived_brand_qualities": [
"accurate",
"easy to use",
"SEO-focused"
],
"competitors": ["SEranking", "ChatGPT Tracker"]
}
```
## Full response with mentions
```json theme={null}
{
"brand": "GenRank",
"domain": "genrank.io",
"geo": "US",
"query": "best chatgpt brand tracker",
"mentions": [
{
"position": 1,
"text_snippet": "GenRank is one of the top tools to track brand visibility in ChatGPT.",
"annotations": [
{
"entity": "GenRank",
"position_start": 0,
"position_end": 7,
"type": "brand"
}
],
"visibility_percent": 72.5,
"perceived_brand_qualities": [
"accurate",
"easy to use",
"SEO-focused"
],
"competitors": ["SEranking", "ChatGPT Tracker"]
}
],
"last_checked": "2025-09-26T10:00:00Z"
}
```
## Interpreting mention data
### Using `position` to assess prominence
A lower `position` value means your brand was mentioned earlier in the response. ChatGPT tends to introduce the most recommended or well-known options first, so a position of `1` is meaningful signal — it suggests your brand is being presented as a leading option for that prompt, not an afterthought.
Tracking position changes over time, particularly after publishing new content or earning coverage, helps you understand whether your optimization efforts are improving your placement within responses.
### Reading `text_snippet` for context
The `text_snippet` field gives you the actual language ChatGPT used when mentioning your brand. This is the most direct signal for understanding framing. A snippet like "GenRank is one of the top tools to track brand visibility in ChatGPT" tells you not just that you were mentioned, but that you were positioned as a leading tool and associated with ChatGPT visibility tracking specifically.
Compare snippets across prompts to understand whether your brand is framed consistently or differently depending on the question context.
### Interpreting `visibility_percent`
Visibility percent is a composite score — it does not measure a single thing but combines several factors into a single number. Use it as a directional signal for trend monitoring rather than as an absolute measure of prominence. A score of 72.5 indicates strong, prominent visibility. Scores below 30 generally reflect brief or peripheral mentions.
Track `visibility_percent` over time across your full prompt set to identify trends. A rising average suggests your brand is becoming more central to AI-generated answers in your category. A sustained drop may signal that a competitor has increased its presence.
### Using `perceived_brand_qualities` for positioning analysis
The qualities in this array are extracted from the language patterns around your brand mention in the ChatGPT response. They reflect how AI currently frames your brand — which may or may not align with how you intend to be positioned.
If you see qualities that do not match your intended positioning (for example, if you want to be known for enterprise scalability but AI describes you as a small business tool), that gap is useful input for your content and GEO strategy.
### Handling empty `mentions` arrays
If the `mentions` array is empty, your brand was not detected in the ChatGPT response for that prompt. This is a prompt gap — a query where you have no AI visibility. Empty mention arrays are valuable data points for competitive analysis: check the `competitors` field in responses from related prompts to see which brands are being recommended in your place.
```python theme={null}
# Filter for prompt gaps (no brand mentions)
gaps = [r for r in responses if len(r["mentions"]) == 0]
print(f"Found {len(gaps)} prompt gaps out of {len(responses)} tracked prompts")
```
# GenRank API: Access Your AI Visibility Data
Source: https://docs.genrank.io/api/overview
Pull brand mentions, visibility scores, sentiment, and competitor data directly from GenRank into your dashboards, apps, and workflows via the GenRank API.
The GenRank API gives you programmatic access to the same brand visibility data that powers the GenRank dashboard. Instead of reading your results in-app, you can pull mentions, rankings, sentiment attributes, and competitor co-occurrence directly into your own tools — internal dashboards, data warehouses, reporting pipelines, or client-facing portals.
The API is available on the **Scale plan** and is set up as a tailored integration. GenRank works with you to define the fields, output format, and delivery schedule that fit your stack before provisioning access.
API access requires the **Scale plan**. If you are on Free, Essential, or Pro, [contact GenRank](https://genrank.io/contact-us/) to discuss upgrading or request an API trial. Scale is custom-priced — book a demo to get a quote.
## What the API provides
The GenRank API surfaces the data collected from running your tracked prompts in ChatGPT every day. Each response in the API reflects a single prompt execution and includes structured fields for brand configuration, mention detail, visibility scoring, perception attributes, and competitor co-occurrence.
Which domains ChatGPT cited in responses to your prompts, including your own domain's citation frequency.
Visibility percent scores and mention rates across all tracked prompts, enabling trend analysis at scale.
Perceived brand qualities attributed to your brand in each response — the language ChatGPT uses to describe you.
Full mention objects including position, text excerpts, entity annotations, and the prompt that triggered the response.
Which competitors appeared in the same response, enabling share-of-voice calculations and competitive benchmarking.
Your brand name, domain, and geo targeting settings — the foundation for all tracked data.
## Who the API is for
The GenRank API is built for teams that want to integrate AI visibility data into existing infrastructure rather than relying solely on the GenRank dashboard. Typical use cases include:
* **Developers and data engineers** building internal visibility dashboards or feeding AI brand data into a data warehouse
* **Digital agencies** delivering white-labeled AI visibility reports to clients from their own platforms
* **Enterprise marketing teams** combining GenRank data with existing analytics stacks (GA4, Looker, Tableau, etc.)
* **Growth and product teams** triggering alerts or automations when brand visibility metrics shift
## How setup works
Because the GenRank API is tailored to each customer's needs, access is provisioned through a brief setup process rather than self-serve signup.
Reach out via [genrank.io/contact-us](https://genrank.io/contact-us/) or complete the API request form on the GenRank API page. Describe your use case, data frequency needs, and any custom fields or outputs you require.
A GenRank team member will meet with you to understand your stack, agree on the fields and output structure you need, and confirm your prompt and geo configuration.
Once the integration is configured, GenRank provisions your API credentials and provides endpoint details specific to your account setup.
Authenticate your requests and begin retrieving brand visibility data programmatically. GenRank remains available for ongoing support as your needs evolve.
## Sample API response
The following example shows the structure of a typical GenRank API response for a single tracked prompt. Field availability may vary based on your configured outputs.
```json theme={null}
{
"brand": "GenRank",
"domain": "genrank.io",
"geo": "US",
"query": "best chatgpt brand tracker",
"mentions": [
{
"position": 1,
"text_snippet": "GenRank is one of the top tools to track brand visibility in ChatGPT.",
"annotations": [
{
"entity": "GenRank",
"position_start": 0,
"position_end": 7,
"type": "brand"
}
],
"visibility_percent": 72.5,
"perceived_brand_qualities": [
"accurate",
"easy to use",
"SEO-focused"
],
"competitors": ["SEranking", "ChatGPT Tracker"]
}
],
"last_checked": "2025-09-26T10:00:00Z"
}
```
## Explore the API reference
How API credentials work, how to authenticate requests, and how to keep your keys secure.
The brand configuration fields — name, domain, and geo targeting — that anchor every API response.
Full reference for the mentions array, including position, text snippets, annotations, and visibility scores.
How competitor data surfaces in API responses and how to use it for share-of-voice analysis.
# GenRank core concepts and terminology
Source: https://docs.genrank.io/core-concepts
Understand prompts, entities, GEO, share of voice, and daily refresh — the key ideas behind how GenRank measures brand visibility in ChatGPT.
GenRank introduces concepts that may be unfamiliar if you are coming from a traditional SEO background. This page explains the most important terms and the ideas behind them. Understanding these concepts will help you interpret your data accurately and make better decisions about where to focus your efforts.
In GenRank, a **prompt** is the exact question or instruction sent to ChatGPT on your behalf. It is the unit of measurement everything else is built around.
A prompt is not a keyword. Keywords are short, fragmented search terms designed for search engine ranking algorithms — "project management software," "best CRM." A prompt is a natural, conversational question that reflects how a person actually talks to an AI tool:
> "What is the best project management software for a startup with a remote team?"
This distinction matters because AI models respond to natural language, not keyword strings. A prompt written to reflect real conversational behavior produces data that accurately reflects how your brand appears in genuine user interactions.
Every metric in GenRank — brand visibility, competitor share, perception — is calculated from responses to specific prompts you have defined and chosen to track. The quality and relevance of your prompt set directly determines the quality of your data.
Prompts can be categorized by intent:
* **Discovery** — "What tools exist for managing social media?"
* **Comparison** — "What is the difference between Notion and Asana?"
* **Transactional** — "Which accounting software should I choose for a freelance business?"
* **Branded** — "Tell me about \[your brand name] and what it does."
These three acronyms describe different optimization strategies for different discovery channels.
**SEO (Search Engine Optimization)** is the practice of improving your content and technical infrastructure to rank higher in traditional search engines like Google. SEO measures rankings, organic traffic, and click-through rates. Success means appearing in a list of ranked links.
**GEO (Generative Engine Optimization)** is the practice of improving how AI tools like ChatGPT represent and recommend your brand. GEO measures brand mentions, citations, share of voice, and entity clarity within AI-generated responses. Success means being included — accurately and favorably — in AI answers to queries relevant to your category.
**AEO (Answer Engine Optimization)** targets a middle layer: featured snippets, AI Overviews in Google, and voice search results. AEO operates within traditional search infrastructure. GEO operates in standalone AI tools that function independently from search engines.
| | SEO | AEO | GEO |
| ---------------- | ----------------- | --------------------------------- | ------------------------------- |
| **Platform** | Google, Bing | Google AI Overviews, voice search | ChatGPT, Claude, Perplexity |
| **Metric** | Rankings, clicks | Featured snippet presence | Brand mentions, share of voice |
| **Content goal** | Rank for keywords | Answer direct questions concisely | Become a citation-worthy source |
GEO does not replace SEO — both serve different purposes and share some foundations, including content quality, authority signals, and technical accessibility. GenRank focuses specifically on GEO measurement for ChatGPT.
GenRank currently tracks ChatGPT, which holds the largest share of AI tool usage globally. Support for tracking additional LLMs is on the roadmap.
**Brand visibility** is a measure of how often your brand is mentioned in ChatGPT responses across your tracked prompt set.
When GenRank runs your prompts daily, it captures the full text of each response and checks whether your brand name appears. Visibility is expressed as a percentage — the proportion of your tracked prompts where a mention was detected.
A prompt-level view shows exactly which individual prompts include your brand and which do not. Prompts where your brand is absent are called **prompt gaps** — they represent queries where other brands are being recommended instead of yours.
Brand visibility is the foundational metric in GenRank. All paid plans include it. The Free plan provides brand visibility tracking for up to 10 prompts.
**Share of voice** measures your brand's mention frequency relative to your tracked competitors across the same set of prompts.
If five brands are mentioned across your tracked prompt set and your brand accounts for 30% of all mentions, your share of voice is 30%. A competitor with 50% of mentions is being recommended more consistently by ChatGPT for queries in your category.
Share of voice gives you a competitive frame for your visibility data. A high absolute visibility score (your brand mentioned in many prompts) can still represent a weak position if competitors are present in nearly every response.
Competitor tracking is available on the Essential plan and above.
In the context of AI language models, an **entity** is a distinct, named concept the model can recognize and resolve — a company, product, person, or place. When ChatGPT processes a prompt about your brand, it needs to identify your brand as a specific, unambiguous entity with consistent properties.
**Entity clarity** is a measure of how consistently and accurately ChatGPT resolves your brand. If your brand name appears in different forms across your content — abbreviated, misspelled, or easily confused with another brand — the model may associate your brand with fragmented or contradictory information. This weakens how reliably you are recognized and recommended.
Strong entity clarity means:
* Your brand name appears in a consistent, canonical form across your content and external sources.
* ChatGPT resolves your brand to the correct entity without conflating it with competitors or similar names.
* Your brand is referenced clearly and repeatedly across AI responses.
GenRank's Entity Clarity tool monitors naming consistency, variation frequency, and fragmentation across captured responses. It is available on the Pro plan.
A **project** in GenRank is a tracking configuration for a single brand and domain. Each project has its own prompt set, competitor list, and visibility data.
You use multiple projects when you need to track separate brands — for example, if you manage a portfolio of clients or run multiple product brands under one account.
The number of projects you can create depends on your plan:
* Free and Essential: 1 project
* Pro: 3 projects
* Scale: custom
All prompts, competitor data, and optimization insights are scoped to the project they belong to. Switching between projects gives you separate dashboards for each brand.
A **prompt slot** is one tracked prompt in your project. Your plan determines the maximum number of prompt slots you can use.
Every day, GenRank runs each prompt in your slot allocation through ChatGPT and captures the response. The number of prompt slots directly limits how many distinct queries you can monitor.
Plan limits:
* Free: 10 prompt slots
* Essential: 25 prompt slots
* Pro: 150 prompt slots
* Scale: custom
Prompt slots are shared across all active prompts in a project. If you are on the Pro plan with 3 projects, you have 150 slots total to distribute across those projects.
Start with fewer, high-intent prompts and expand as you learn which prompt categories produce the most useful signal. A focused set of well-chosen prompts delivers more actionable data than a large set of loosely relevant ones.
The **daily refresh** is the cycle in which GenRank re-runs all your tracked prompts through ChatGPT and captures fresh responses.
Because ChatGPT generates responses dynamically — and because it can access real-time web content — the brands mentioned in responses to any given prompt can change over time. A daily refresh ensures your visibility data reflects current AI behavior rather than a static snapshot.
Key implications of the daily refresh:
* Visibility scores are updated every 24 hours.
* Changes caused by new content you publish, PR coverage, or competitor activity become detectable within days.
* Historical data accumulates over time, allowing you to identify trends, validate optimization efforts, and detect when visibility shifts.
All GenRank plans include daily refresh.
When ChatGPT performs a web search before generating a response, it may cite external sources — domains and pages it retrieved to inform its answer. GenRank logs these citations.
**Sources** tracking shows you:
* Which external domains are cited in responses to your tracked prompts.
* How frequently each domain appears.
* Whether your own domain is being cited.
* Which domains support competitor mentions and which support yours.
Citations matter because they reflect the content that AI considers credible and relevant for queries in your category. If a competitor's domain is consistently cited while yours is not, that is a signal about content authority and retrieval alignment — not just a visibility gap.
Sources data is available on the Essential plan and above. The Content Retrieval tool in Optimization (Pro plan) provides deeper analysis of what elements make cited pages retrievable.
**Brand perception** in GenRank refers to the recurring qualities, themes, and attributes that ChatGPT associates with your brand across responses.
AI language models attach patterns to named entities based on the content they have processed and retrieved. These patterns shape how a brand is described — reliably, at scale — to anyone who asks. GenRank extracts and quantifies those attributes so you can see how your brand is actually positioned in AI-generated descriptions, not just whether it is mentioned.
Examples of perception attributes might include terms like "affordable," "enterprise-grade," "easy to use," "for small businesses," or "complex setup." You can compare how your brand's attributes differ from competitors' in the same responses.
Brand perception is available on the Essential plan and above.
## Putting the concepts together
A typical GenRank workflow moves through three stages that build on each other:
1. **Define prompts** (Prompt Research) — Identify which queries your customers ask ChatGPT in your category. These become the inputs for all measurement.
2. **Measure performance** (Response Tracking) — Track your brand visibility, share of voice, competitor positions, perception, and source citations across those prompts, updated daily.
3. **Improve results** (Optimization) — Use content retrieval analysis and entity clarity monitoring to understand the structural reasons behind your visibility and make targeted changes.
Follow the step-by-step guide to set up your first project and prompts.
Learn how to build and manage your prompt library.
# What is GenRank and how does it work?
Source: https://docs.genrank.io/introduction
GenRank is a GEO platform that tracks how your brand appears in ChatGPT responses — measuring mentions, competitors, and perception daily.
GenRank is a generative engine optimization (GEO) platform built to make your brand's presence in AI-generated answers measurable. Every day, millions of people ask ChatGPT questions that lead to brand recommendations. GenRank tells you whether your brand appears in those answers, how it is described, who else is mentioned, and what you can do to improve your position.
## Why AI visibility matters
When someone asks ChatGPT "what's the best project management tool for remote teams?" or "which accounting software should a small business use?", the response shapes purchasing decisions without producing a list of ranked links. There is no page two. There is no paid placement. There are only the brands ChatGPT chooses to mention — and those it ignores.
Traditional SEO tools were built to track your position in Google's search results. They are not designed to answer the question: "Does ChatGPT recommend my brand, and if so, how?" GenRank was built specifically for that question.
GenRank currently tracks visibility in **ChatGPT**, which accounts for roughly 69% of AI tool usage worldwide. Support for additional LLMs is planned.
## How GenRank differs from traditional SEO
Search engine optimization and generative engine optimization share some foundations — quality content, credibility signals, clear site structure — but they measure entirely different things.
| | Traditional SEO | GEO (GenRank) |
| ------------------ | -------------------------------- | -------------------------------------------- |
| **Platform** | Google, Bing | ChatGPT |
| **Success metric** | Rankings and clicks | Mentions and citations |
| **Competition** | Finite positions (1st, 2nd, 3rd) | Share of voice across responses |
| **Content goal** | Rank for keywords | Become a citation-worthy authority |
| **Measurement** | Traffic and rankings | Brand visibility, perception, entity clarity |
GEO does not replace SEO. It adds a complementary measurement layer for an increasingly important discovery channel.
## The three pillars of GenRank
GenRank is organized around three interconnected workflows. You start by identifying the right prompts to track, then monitor your brand performance across those prompts, and finally use diagnostic tools to understand and improve your results.
Discover the questions your customers ask ChatGPT. Build your tracking library from Google Search Console data, URL scans, or manual input.
Monitor brand mentions, competitor visibility, brand perception, and source citations — updated daily from real ChatGPT responses.
Understand why you appear or don't. Analyze content retrieval alignment, entity clarity, and the sentiment signals that influence AI recommendations.
### Prompt Research
Before you can measure anything, you need to define which prompts matter. A prompt in GenRank is the actual question or query that gets sent to ChatGPT — for example, "best CRM software for a sales team under 20 people." The prompts you track determine everything that follows.
GenRank gives you three ways to build your prompt library:
* **Prompt Manager** — Create, import, tag, and organize prompts manually. Every prompt can be categorized by intent: discovery, comparison, transactional, or branded.
* **Page Scan** — Enter a URL from your site and GenRank generates relevant prompts a user might ask ChatGPT about that page, product, or service.
* **Google Search Console integration** — Pull long-form, question-based queries from your existing search data and convert them into prompts ready for AI tracking.
### Response Tracking
Once your prompts are set up, GenRank runs them in ChatGPT every day and captures the full responses. The data is structured into four reporting areas:
* **Brand Visibility** — How often your brand is mentioned across tracked prompts, your share of voice relative to competitors, and which prompts you do and do not appear in.
* **Competitors** — Which competing brands appear in the same responses, which prompts they dominate, and where you have visibility gaps.
* **Brand Perception** — The recurring qualities, themes, and attributes that ChatGPT associates with your brand across responses.
* **Sources** — Which external domains are cited in responses about your category, and how frequently your own domain appears.
### Optimization
Tracking tells you what is happening. Optimization tells you why, and what to change. GenRank's optimization tools translate raw AI responses into concrete insights:
* **Content Retrieval** — Analyzes which content elements appear in pages that get cited, and how closely your content aligns with what AI retrieves for each prompt.
* **Entity Clarity** — Monitors how consistently ChatGPT resolves your brand as a single, unambiguous entity. Naming inconsistencies and fragmentation weaken recognition.
* **Sentiment Radar** — Tracks how your brand is described across non-owned sources — reviews, news, social platforms — to surface the reputation signals that influence AI recommendations.
## Who GenRank is for
GenRank is used by brand managers, SEO professionals, digital agencies, and marketing teams who need to understand and grow their brand's presence in AI-generated discovery. If your customers use ChatGPT to research products or services in your category, GenRank gives you the data to measure and act on that visibility.
Create your account, set up a project, and add your first prompts.
Understand the terminology and ideas behind how GenRank works.
# Content Retrieval
Source: https://docs.genrank.io/optimization/content-retrieval
Discover what content elements AI selects for a prompt, compare your pages with vector similarity analysis, and close gaps that keep you out of responses.
When ChatGPT answers a question, it selects content based on structural alignment — not keyword density. Content Retrieval shows you exactly what that alignment looks like for any prompt you track: what elements AI consistently picks up, how your own content compares at the chunk level, and which claims your competitors have locked down. The result is a concrete, page-checkable list of changes you can make before the next response cycle runs.
## Content requirements
Content requirements are the structured output of GenRank's analysis of real ChatGPT responses for a given prompt. Rather than showing you what AI said, they translate those responses into a checklist your content team can act on.
GenRank analyzes a batch of real LLM responses for a prompt and extracts recurring patterns across them:
* **Answer formats** — does AI answer with a numbered list, a comparison table, a definition-first paragraph, a step-by-step guide?
* **Claims** — which factual assertions appear repeatedly, regardless of the source?
* **Definitions** — which concepts does AI consistently define, and how?
* **Comparisons** — which alternatives or competitors does AI routinely place alongside your brand?
* **Attributes** — what properties, features, or qualifiers does AI associate with solutions in this category?
* **Supporting elements** — what kinds of evidence (statistics, use cases, testimonials) does AI draw on to substantiate its answers?
Each of these patterns becomes a page-checkable requirement. If AI consistently includes a specific type of definition when answering a prompt, that's not coincidence — it's a retrieval signal. Your page either satisfies it or it doesn't.
Content requirements update when GenRank re-analyzes response batches. Revisit them periodically, especially after major competitor content changes or shifts in your prompt's mention rate.
## Vector similarity analysis
Content requirements tell you what AI looks for. Vector similarity analysis tells you whether your content delivers it — and exactly where it falls short.
GenRank simulates how LLMs embed your content and compares those embeddings against the real prompts you track. This analysis runs at the **chunk level**, meaning your page is broken into sections (typically by heading or paragraph block) and each chunk is scored independently for semantic alignment with the prompt.
### What you see
For each URL you analyze, you get a visual breakdown of chunk-level similarity scores:
* **High-alignment chunks** — sections that closely match the prompt's embedding space. These are sections AI is likely to retrieve when generating a response.
* **Low-alignment chunks** — sections that are semantically distant from the prompt. These may be well-written but are structurally irrelevant to what AI is trying to answer.
* **Missing coverage zones** — prompts topics or angles that none of your chunks address at all.
### How to interpret the results
A low overall similarity score means your page is unlikely to be retrieved for that prompt, even if it ranks well on Google. A high overall score with several weak chunks suggests your content is partially aligned — AI may retrieve part of your page but miss the most authoritative sections.
Look for patterns across multiple prompts. If your introductory chunks consistently score high but your supporting-evidence sections score low, you likely need to restructure how you present proof and specifics.
Chunk-level analysis surfaces gaps that keyword tools can't detect. A section that uses the right words but structures them as a promotional narrative rather than a direct answer will score low — because it doesn't match the embedding pattern of the content AI selects.
## Dominant claims
Dominant claims are the recurring assertions that AI makes when answering prompts in your market category — and the brands those assertions are attributed to.
GenRank extracts these claims from response batches and maps them to specific brands. A claim like "the most accurate AI citation tracker" appearing repeatedly and attributed to a competitor tells you that competitor has effectively occupied that position in AI's understanding of your category.
Use dominant claims data to answer three questions:
1. **Which claims define your category?** If a claim appears in the majority of AI responses for a prompt, it's load-bearing for that topic. Not owning it means ceding positioning to whoever does.
2. **Who currently controls each claim?** Attribution frequency shows where authority is concentrated. A claim attributed to one brand across most responses is entrenched.
3. **Are any valuable claims unattributed?** Claims that appear frequently but aren't consistently attributed represent positioning opportunities — no brand has yet established clear authority.
## Practical workflow
Use this workflow when you want to improve your retrieval rate for a specific prompt.
Open **Optimization → Content Retrieval** and choose a prompt from your tracked list. Prioritize prompts where your mention rate is low or where a competitor appears significantly more often than you do.
Read through the requirements GenRank has extracted from real AI responses for that prompt. Note which answer formats, claims, and supporting elements appear most frequently — these are the highest-signal requirements.
Enter the URL of the page you intend to rank for this prompt. Review the chunk-level similarity scores to see which sections align well and which don't.
Cross-reference low-scoring chunks with the content requirements list. If a high-frequency requirement (for example, a direct comparison table) has no corresponding chunk on your page, that's a confirmed gap.
Revise the page to address the gaps. Add the missing content elements, restructure weak sections so each chunk directly addresses a distinct aspect of the prompt, and ensure supporting evidence is concrete and specific rather than general.
After publishing the updated page, re-run the vector similarity analysis to confirm your chunk scores have improved. Monitor your mention rate for that prompt over the following tracking cycles.
## Frequently asked questions
Traditional SEO focuses on keyword presence and technical ranking signals. Content Retrieval focuses on semantic alignment — whether your content structurally matches what AI selects when generating a response. A page can rank on page one of Google and still score poorly for AI retrieval if its structure doesn't match the patterns AI favors for that prompt type.
GenRank re-analyzes response batches periodically and whenever your prompt's tracking data shows a significant shift. You can also trigger a manual re-analysis from the Content Retrieval dashboard.
Yes. You can run vector similarity analysis on any publicly accessible URL, including competitor pages. This lets you understand why a competitor's page is being retrieved over yours for a specific prompt.
GenRank splits pages by major heading sections and logical paragraph blocks. A typical blog post might produce 8–20 chunks. Very long pages may produce more. The chunking logic mirrors how LLMs break content during the retrieval phase of their inference pipeline.
# Entity Clarity
Source: https://docs.genrank.io/optimization/entity-clarity
Measure how consistently AI recognizes your brand as one entity. Detect naming variations, alias drift, and fragmentation before they erode your visibility.
Language models don't retrieve your brand the way a search engine indexes a domain. They resolve it as an entity — a named concept that can be matched, attributed, and referenced across training data and live responses. If your brand name appears in inconsistent forms across the web, AI systems struggle to treat all those references as the same entity. Mentions scatter. Citations weaken. Entity Clarity measures this fragmentation and gives you the data to fix it.
## Why entity consistency matters
When AI generates a response that mentions your brand, it's drawing on a learned representation built from thousands of references across the web, structured databases, and its training corpus. The more consistent those references are — in spelling, structure, and context — the stronger and more confident that representation becomes.
Inconsistency works against you in three ways:
* **Split authority** — if "GenRank," "Gen Rank," and "Genrank.io" all appear as separate-seeming entities, the authority signals attached to each are diluted rather than compounded.
* **Attribution drift** — AI may correctly know a fact about your brand but fail to attribute it to you because the name in context doesn't cleanly match the entity it has the strongest representation for.
* **Recognition gaps** — in some cases, a fragmented entity name simply doesn't match the threshold for confident entity resolution, and AI omits your brand entirely rather than risking an incorrect attribution.
## What entity fragmentation looks like
Fragmentation isn't limited to obvious misspellings. Common patterns include:
* **Capitalization variations** — "GenRank" vs. "Genrank" vs. "GENRANK"
* **Spacing variations** — "GenRank" vs. "Gen Rank" vs. "Gen-Rank"
* **Domain-as-name** — "GenRank" vs. "GenRank.io" vs. "genrank.io"
* **Alias drift** — informal shorthand ("GR," "the GenRank platform," "GenRank's tool") that accumulates across partner sites, reviews, and press coverage without a clear link back to the canonical name
* **Structural ambiguity** — names that overlap with other brands, common words, or acronyms in adjacent industries
## Defining your canonical entity
The first step in using Entity Clarity is telling GenRank what your brand's canonical entity name is. This is the exact form you want AI models to use — the one that appears on your homepage, in your legal name, and in your structured data.
To set your canonical entity name:
1. Open **Optimization → Entity Clarity**.
2. Click **Edit canonical entity**.
3. Enter the exact name you want AI to use (for example, `GenRank`).
4. Save. GenRank will use this as the baseline for all fragmentation analysis.
Choose the name form that is most consistently used across your highest-authority pages and structured data sources. Changing your canonical entity name after extended monitoring will reset your fragmentation baseline.
## Reading the entity fragmentation report
The fragmentation report shows you how consistently AI references your brand across the responses GenRank has captured for your tracked prompts.
The report surfaces:
* **Fragmentation score** — a composite measure of how often AI uses a form other than your canonical name. Lower is better. A score above 20% typically indicates a meaningful consistency problem.
* **Variation inventory** — a ranked list of every non-canonical name form detected in AI responses, sorted by frequency. This shows you which variants are most prevalent.
* **Alias drift over time** — a timeline view of fragmentation trends. Rising fragmentation after a product rename, rebrand, or press spike is a common and actionable pattern.
* **Context samples** — for each detected variation, you can view the response excerpt where it appeared. This helps you understand whether a variant is appearing in a neutral context (a simple mention) or an attributive context (a claim, comparison, or recommendation).
Fragmentation is measured across all responses GenRank captures for your tracked prompts. The more prompts you track, the more comprehensive the fragmentation picture becomes.
## Steps to improve entity clarity
Reducing fragmentation requires consistent action across the web properties that feed into AI training data and live citations. Work through these steps in order — the earlier items have the highest leverage.
Review every page on your domain and ensure the brand name appears in exactly the canonical form. Check page titles, headings, the About page, footer text, and structured data (`schema.org/Organization`, `og:site_name`). Your own site is the highest-authority source for your entity.
Implement `schema.org/Organization` on your homepage with a consistent `name`, `url`, `logo`, and `sameAs` array pointing to your official social profiles and knowledge base entries. This gives AI systems a machine-readable declaration of your canonical entity and its equivalents.
Identify high-frequency sources of variation in the fragmentation report. Reach out to partner sites, directories, and review platforms where your brand appears in a non-canonical form and request corrections. Prioritize sources that AI responses cite most frequently — these have the greatest impact on entity resolution.
Wikipedia and Wikidata entries are strong entity resolution anchors for LLMs. If your brand has a Wikipedia page, ensure the name in the opening sentence exactly matches your canonical form. If it doesn't have one, consider whether your brand meets notability guidelines. Wikidata entries can be created independently and are referenced by many AI systems.
News coverage is a significant source of entity training signal. When working with journalists or issuing press releases, always use your canonical entity name in the official company name field and first mention. Avoid informal shorthand in official communications.
Return to the fragmentation report after major content pushes, product launches, and press coverage spikes. Entity fragmentation can re-emerge after rebrands or when new external content is published at volume. Treat fragmentation monitoring as an ongoing practice, not a one-time audit.
## Frequently asked questions
Not always directly, but consistently. When AI can't confidently resolve your brand as a single entity, it defaults to the variant with the strongest representation — which may produce fewer overall mentions than a unified entity would. Severe fragmentation can also cause AI to omit your brand from responses where it would otherwise appear.
Ambiguity with another entity is a separate problem from internal fragmentation, but Entity Clarity surfaces it. In these cases, structured data disambiguation — using `sameAs` links to your official profiles and differentiating context in your structured data — is the most effective strategy. GenRank will flag when AI appears to be confusing your brand with another entity.
Changes to your own site and structured data can improve entity resolution within weeks as AI systems that perform live web retrieval update their understanding. Changes to training data have a longer feedback loop — typically tied to model update cycles, which vary by LLM provider.
Only if your brand is universally known by its domain name (for example, if "genrank.io" is how you appear in press coverage and partner sites). In most cases, the brand name without the TLD is the correct canonical form. Avoid including the TLD unless it's genuinely part of your public-facing brand identity.
# Overview
Source: https://docs.genrank.io/optimization/overview
Move beyond tracking. Optimization reveals why your brand appears in ChatGPT and what to change — improving retrieval, entity clarity, and sentiment signals.
Tracking tells you where your brand appears in ChatGPT responses. Optimization tells you why you appear — or why you don't — and what to change. While Response Tracking delivers the score, the Optimization module delivers the diagnosis and the prescription: structural content gaps, entity fragmentation, and the external sentiment signals that shape how AI perceives your brand.
## What's in the Optimization module
The module is organized around three distinct lenses, each targeting a different reason why AI may overlook or misrepresent your brand.
Understand what content elements AI selects when answering a prompt. Compare your pages against those requirements using vector similarity analysis at the chunk level.
Measure how consistently AI models recognize your brand as a single, unambiguous entity. Detect naming variations and alias drift before they weaken your visibility.
Monitor how your brand is described across customer reviews, social platforms, news, and other non-owned sources that feed into AI training data and citations.
## Tracking vs. optimization
These two modules are designed to work together, not independently.
| | Response Tracking | Optimization |
| -------------------- | --------------------------------------------------- | --------------------------------------------------------------------- |
| **Primary question** | Where do I appear? | Why do I appear or not? |
| **Output** | Mention rate, citation count, competitor benchmarks | Content requirements, entity fragmentation score, sentiment breakdown |
| **Use case** | Daily performance monitoring | Diagnosing gaps and planning content changes |
| **Feedback loop** | Measures change after you act | Identifies what actions to take |
Use Tracking to detect shifts in your visibility. Use Optimization to understand what's driving those shifts and how to address them.
## How AI selects content
When ChatGPT answers a prompt, it doesn't retrieve a ranked list of web pages — it selects content based on semantic alignment, entity resolution, and the credibility of cited sources. That means three distinct factors can hold your brand back:
1. **Retrieval gaps** — your content doesn't structurally match what AI is looking for when answering a specific type of question.
2. **Entity fragmentation** — AI can't consistently identify your brand as one clear entity, so mentions scatter across variations.
3. **Sentiment signals** — the external web narrative around your brand is mixed or negative, reducing AI's confidence in citing you.
Each Optimization tool addresses one of these factors directly.
Content Retrieval, Entity Clarity, and Sentiment Radar are available on the **Pro plan and above**. Upgrade your plan in **Account → Billing** to unlock the full Optimization module.
# Sentiment Radar
Source: https://docs.genrank.io/optimization/sentiment-radar
Monitor how your brand is described across reviews, social platforms, news, and trust signals — the sentiment that shapes AI perception and citation behavior.
AI models don't only pull from content you publish. They draw on the broader web narrative around your brand — the reviews customers leave, the coverage journalists write, the discussions happening on social platforms. Sentiment Radar monitors that narrative across non-owned sources, groups it by reputation channel, and surfaces the tone patterns that influence how AI understands and describes your brand. It gives you visibility into the external signals you didn't write but can still influence.
## What Sentiment Radar monitors
Sentiment Radar focuses exclusively on sources you don't control — the external web mentions that feed into AI training data and inform the citations AI systems produce when recommending or describing brands in your category.
These sources matter because AI doesn't evaluate your brand only by what your website says. When generating a response, an LLM's understanding is shaped by the aggregate signal across all sources it has processed. A brand with strong owned content but overwhelmingly negative external mentions will often be described with qualifications, caveats, or omitted in favor of competitors with cleaner reputations.
## The five reputation channels
Sentiment Radar organizes external mentions into five channels, each reflecting a distinct audience and type of influence signal.
Mentions from product review platforms, app stores, G2, Capterra, Trustpilot, and similar sites. These carry high weight because they represent direct buyer experience and are frequently cited by AI when answering evaluation or comparison prompts.
Mentions from Glassdoor, Indeed, Blind, and similar employer review platforms. While not directly relevant to product decisions, AI uses these as legitimacy and stability signals when assessing brand credibility.
Mentions from LinkedIn, Reddit, X (formerly Twitter), and other public social channels. These reflect real-time community sentiment and often carry emerging narratives before they appear in formal coverage.
Mentions from business directories, accreditation bodies, industry associations, and fact-checking sources. These are strong anchors for AI entity confidence — they signal that your brand is recognized by authoritative third parties.
Mentions from news publications and industry media. News citations are among the highest-authority external signals AI systems use. A positive news mention contributes significantly to citation likelihood; a critical one can introduce persistent qualifications into AI responses.
## Interpreting your sentiment data
The Sentiment Radar dashboard shows you a channel-by-channel breakdown of sentiment polarity (positive, neutral, negative) alongside the volume of mentions detected in each channel over your selected time range.
When reading the report, focus on these areas:
**Where positive sentiment is strongest.** Channels showing high positive mention volume are working for you. They're contributing favorable signals to the external narrative AI draws on. Identify what's driving positive mentions in those channels and look for ways to replicate that on channels where your score is weaker.
**Where criticism concentrates.** A channel with a disproportionate share of negative sentiment is a liability. AI systems that cite sources in that channel will absorb that tone into their understanding of your brand. A high volume of critical customer reviews, for example, can cause AI to add unprompted caveats when recommending your product.
**Which narratives are spreading.** The trend view shows whether sentiment in a channel is improving, stable, or deteriorating over time. A sudden spike in negative social mentions often precedes a shift in AI tone within days to weeks, as AI systems update their retrieval indexes.
**Gaps in trust signals.** If the trust and legitimacy channel shows low mention volume, your brand lacks the third-party validation anchors that AI systems use to establish entity credibility. This is a common but fixable gap.
Sentiment Radar captures publicly accessible external mentions. It does not monitor private or authenticated content, internal communications, or paywalled sources.
## The connection between external sentiment and AI perception
AI systems are citation-aware. When a language model generates a response that describes your brand, it's drawing on a learned representation built from everything it has processed about you — including external sources with strong positive or negative tone signals.
This means that external sentiment affects AI perception through two mechanisms:
1. **Training data influence** — the tone of external content included in an LLM's training corpus shapes how the model describes your brand by default, without any live retrieval happening.
2. **Live retrieval influence** — for AI systems that perform real-time web retrieval (like ChatGPT with Browse enabled), negative or credibility-undermining sources can be pulled into responses directly, sometimes as explicit citations.
Neither mechanism is fully within your control, but both are influenced by the quality and tone of your external web presence. Improving sentiment on high-authority channels reduces the likelihood that AI will characterize your brand negatively or add unsolicited qualifications to its recommendations.
Focus first on improving sentiment in the channels AI systems cite most often in your category. Check your **Response Tracking → Sources** report to see which external domains appear most frequently in ChatGPT responses about your topic area. If G2 or Trustpilot pages appear there regularly, those review channels have outsized influence on AI perception for your market.
## Acting on sentiment signals
Use the following approaches to address the issues Sentiment Radar surfaces.
**For negative customer review signals:** Respond promptly and constructively to negative reviews on high-visibility platforms. AI systems index tone patterns, not just individual reviews — a consistent pattern of resolved complaints signals responsiveness more strongly than isolated positive reviews.
**For weak trust and legitimacy signals:** Apply to relevant industry directories, accreditation programs, and business associations that are authoritative in your category. Request reviews or mentions from partners and analysts who publish on indexed platforms. Each verified third-party mention strengthens AI's entity confidence in your brand.
**For negative or thin news coverage:** Consider a proactive PR effort — data studies, expert commentary, or product announcements that give journalists citable, neutral-to-positive material. Even a single high-authority news mention can shift the tone balance in AI responses for that topic area.
**For emerging social narratives:** Monitor the sentiment trend view weekly. When a new negative narrative begins spreading on social platforms, address it in public-facing content before it has time to propagate to higher-authority channels or get indexed into AI retrieval systems.
Sentiment Radar monitors and reports on external signals. Improving those signals requires action in the channels themselves — publishing responses, generating new coverage, or earning new trust-signal mentions. GenRank tracks the results of those efforts automatically as new data is captured.
# Prompt Suggestions
Source: https://docs.genrank.io/prompt-research/page-scan
Enter any URL and GenRank generates the prompts users are likely to ask ChatGPT about that page, product, or brand — no manual guesswork required.
Page Scan turns your existing website content into a structured prompt library. Instead of manually inventing questions your customers might ask, you provide a URL and GenRank analyzes the page to generate a list of realistic prompts users would type into ChatGPT to find, evaluate, or learn about what that page represents. It bridges the gap between the content you have already created and the conversational behavior happening in AI.
## What Page Scan does
When you submit a URL, GenRank reads the page content — the headlines, product descriptions, features, use cases, and positioning language — and generates prompts that reflect natural user intent around that content. The output is a list of ready-to-review suggestions organized by prompt type, which you can selectively add to your Prompt Manager.
This is especially useful when:
* You have product or feature pages that represent specific use cases but haven't yet mapped those to AI queries
* You launch a new landing page and want immediate prompt coverage without starting from scratch
* You publish a blog post and want to understand what AI visibility that content could support
* You want to audit how well your current content aligns with the questions people actually ask
Page Scan works on any publicly accessible URL. If a page is behind a login or restricted by robots.txt, the scan may return limited results or be unable to complete.
## How to use Page Scan
Open **Prompt Research** in the left sidebar and select **Page Scan**.
Paste the full URL of the page you want to analyze into the input field. This can be your homepage, a product page, a feature landing page, a blog post, or any public page on your site.
Click **Scan**. GenRank fetches the page content and runs it through its prompt generation model. Most scans complete within 30 to 60 seconds depending on page length and complexity.
The results panel displays a list of generated prompts grouped by intent type — discovery, comparison, transactional, and branded. Each prompt appears as a standalone suggestion you can preview before deciding whether to track it.
Check the prompts that are relevant to your tracking goals. You can select individual prompts or choose all prompts in a category at once. Deselect any that are too generic, inaccurate, or outside the scope of what you want to monitor.
Click **Add Selected**. The chosen prompts are transferred directly to your Prompt Manager with their intent categories pre-assigned. From there you can adjust categories, enable web retrieval, or edit the prompt text before they go live.
## When to run a Page Scan
Page Scan is not a one-time setup step — it is a recurring input as your content grows and evolves.
Run a scan on new landing pages immediately after publication. Capture the prompts users will ask about your new offering before your competitors establish visibility in those queries.
Informational content often maps to discovery-stage queries. Scan articles, guides, and comparison pages to find the conversational prompts your content can support.
Your homepage and top-level product pages represent your broadest positioning. Scanning them surfaces the core queries that define how your category and brand are perceived in AI.
You can scan publicly accessible competitor URLs to identify which prompts their content is optimized to address — and compare that against your own coverage.
## Understanding the results
Page Scan generates prompts based on what your page represents, not what it ranks for. The output reflects conversational intent derived from your content, so the quality of the suggestions depends on how clearly your page communicates its purpose.
If a page scan returns prompts that seem off-target, consider whether the page itself clearly articulates:
* What the product or content is
* Who it is for
* What problem it solves or question it answers
* How it compares to alternatives
Pages with vague or sparse copy tend to produce generic prompt suggestions. Pages with specific use-case language, clear audience targeting, and concrete problem framing produce more actionable prompt sets.
Use Page Scan output as a signal about your content clarity. If the generated prompts don't reflect the queries you want to own, that may indicate your page needs sharper positioning — both for AI visibility and for the humans reading it.
## Adding scanned prompts to your library
After selecting prompts from a scan, they appear in your Prompt Manager as inactive by default, giving you a chance to review and configure them before tracking begins. Activate the prompts you want GenRank to run, and they will be included in the next scheduled tracking cycle.
Prompts added through Page Scan count toward your plan's prompt slot limit. If you are approaching your limit, review your existing library and deactivate any prompts that are no longer relevant before adding new ones from a scan.
# Prompt Manager
Source: https://docs.genrank.io/prompt-research/prompt-manager
Create, tag, bulk-import, and monitor every prompt in one place. Control web retrieval mode per prompt and keep your tracking library organized at scale.
The Prompt Manager is your central workspace for every prompt GenRank tracks. You define the prompts here, assign them intent categories, toggle web retrieval behavior, and control which prompts are actively being monitored. Every visibility metric, competitor comparison, and trend chart in GenRank flows from the prompts you manage in this view.
## Adding your first prompt
Navigate to **Prompt Research** in the left sidebar, then select **Prompt Manager**. You will see your existing prompt library, or an empty state if this is your first session.
Select **Add Prompt** in the top-right corner of the library view. A prompt creation panel opens on the right side of the screen.
Enter the full prompt text exactly as a user would type it into ChatGPT. Use natural, conversational language — not keyword fragments. For example, write "what is the best accounting software for freelancers" rather than "accounting software freelancers."
Choose the category that best describes the intent behind this query. See the [intent categories section](#intent-categories) below for guidance on each option.
Decide whether this prompt should force ChatGPT's web search mode when GenRank runs it. Enable this if you want to test visibility in retrieval-augmented responses. See the [web retrieval section](#web-retrieval-control) below for more detail.
Click **Save**. The prompt is added to your library and begins tracking in the next scheduled run. Newly added prompts are activated by default — you can deactivate any prompt at any time without losing existing data.
## Intent categories
Every prompt in GenRank belongs to an intent category. Categorizing prompts lets you analyze visibility by funnel stage — so you can see where you dominate and where you are underrepresented.
Discovery prompts reflect early-stage research behavior. Users are exploring a category, not yet evaluating specific options. These are typically "what is the best X," "top X tools," or "how do I solve Y" type queries.
**Examples:**
* "what is the best email marketing platform for small businesses"
* "top project management tools for startups"
* "how do companies manage customer support at scale"
Discovery prompts tend to have high volume and are where brand awareness in AI is first established.
Comparison prompts signal that a user is evaluating options. They often name specific products or pairings directly. These are "X versus Y," "alternatives to X," or "which is better" queries.
**Examples:**
* "HubSpot vs Salesforce for a mid-market SaaS company"
* "best alternatives to Notion for knowledge management"
* "which is better for e-commerce, Shopify or WooCommerce"
Comparison prompts are critical for competitive intelligence — they reveal whether ChatGPT favors you or a competitor when users are ready to choose.
Transactional prompts reflect high purchase intent. Users are looking for a specific solution for a specific need. These queries often include buyer-context details like company size, budget, use case, or role.
**Examples:**
* "best CRM for a B2B SaaS startup with a small sales team"
* "affordable HR software for companies with 50 to 200 employees"
* "what accounting tool should a freelance designer use"
Transactional prompts typically have the highest conversion relevance and the most direct impact on buying decisions.
Branded prompts include your brand name directly. They capture how ChatGPT describes, positions, and represents your brand when users ask about it specifically.
**Examples:**
* "what is \[Your Brand]"
* "is \[Your Brand] good for enterprise teams"
* "how does \[Your Brand] compare to its competitors"
Tracking branded prompts is essential for understanding your brand entity clarity — how accurately and consistently ChatGPT represents what you do.
Custom is a free-form category for any prompt that doesn't fit the standard intent types. Use it for industry-specific queries, audience-segment prompts, or experimental tracking that falls outside the standard funnel structure.
## Web retrieval control
Each prompt has an optional web retrieval toggle. When enabled, GenRank forces ChatGPT to use its web search mode when running that specific prompt. When disabled, the prompt runs without web retrieval, using only ChatGPT's base model knowledge.
This distinction matters because AI responses can differ significantly depending on whether external sources are retrieved. A brand that appears in ChatGPT's training data may be described differently than one that is primarily known through recent web content.
Use cases for enabling web retrieval:
* Testing whether your recent content, announcements, or product updates are being picked up
* Understanding how third-party sources like review sites or media coverage shape your brand's representation
* Comparing your visibility with and without retrieval to identify where web presence is a factor
Run a paired test: add the same prompt twice, once with web retrieval enabled and once without. Compare the responses in Response Tracking to understand how retrieval changes your brand's positioning.
## Bulk import
If you have a large set of prompts — from keyword research, customer interviews, competitor analysis, or export from another tool — you can add them all at once using bulk import.
Format your prompts as a plain text list, one prompt per line. You do not need to include categories or metadata at this stage — those can be assigned after import.
In Prompt Manager, select **Import** and choose **Bulk Add**. Paste your prompt list directly into the input field or upload a `.txt` or `.csv` file.
GenRank displays a preview of all imported prompts before adding them. Review for duplicates or prompts that need editing, then assign intent categories. You can apply a category to all prompts in the batch or categorize them individually.
Click **Import** to add all reviewed prompts to your library. Prompts that would exceed your plan's slot limit are flagged — you can upgrade your plan or deactivate existing prompts to free up slots.
Bulk import does not check for exact duplicates across your existing library. Review your current prompts before importing a large batch to avoid tracking the same query multiple times and consuming unnecessary prompt slots.
## Prompt selection strategy
**Start narrow, then expand.** Begin with 5–10 high-signal prompts that directly reflect how your category is discovered and compared. Establish a baseline, review your first two weeks of data, and use those results to identify gaps before scaling your prompt set.
When building your prompt library, consider the following principles:
**Write prompts the way real users ask them.** "What's the best invoicing software for a solo consultant" captures real intent. "Invoicing software consultant best" does not. ChatGPT responds to conversational input, and your prompts should reflect that.
**Cover the full funnel, not just the bottom.** It's tempting to focus only on high-intent transactional prompts, but discovery and comparison queries often have higher volume and shape brand perception before a purchase decision is made.
**Include competitor-framing prompts.** Queries like "alternatives to \[Competitor]" or "\[Competitor] vs \[Your Brand]" reveal whether ChatGPT positions you favorably when users are actively evaluating options.
**Revisit your prompt set regularly.** Prompt landscapes evolve as products change, new competitors emerge, and user behavior shifts. A prompt library that was accurate six months ago may no longer reflect current conversations in your category.
**Use Page Scan and Search Console as inputs.** Rather than inventing prompts manually, derive them from signals you already own — your existing content and your real search data. See [Page Scan](/prompt-research/page-scan) and [Search Console](/prompt-research/search-console) for details.
# Queries
Source: https://docs.genrank.io/prompt-research/search-console
Connect Google Search Console to surface multi-word, question-based queries your site ranks for and convert real search demand into tracked AI prompts.
The Google Search Console (GSC) integration lets you build your AI prompt library on top of demand signals you already own. Your GSC data contains the actual queries people type into Google to find your site — and a significant portion of those queries are long, multi-word, and conversational. These are precisely the kinds of questions that are migrating from traditional search to AI. By surfacing them in GenRank, you convert proven search demand into AI tracking candidates without starting from guesswork.
## Why this matters
The gap between traditional search queries and AI prompts is smaller than most people expect. When someone types "best project management software for remote teams with time tracking" into Google, they are one small step away from asking ChatGPT the same thing in sentence form. That query is already in your GSC data if your site ranks for it. GenRank identifies which of your existing queries have this conversational structure and surfaces them as ready-to-use prompts.
Building your prompt library from search data gives you three advantages:
1. **Proven demand**: these queries already drive real traffic to your site, which means real users care about them
2. **Category relevance**: your site ranks for them because they match your content — so they reflect genuine positioning opportunities
3. **Migration signal**: long, question-based queries are the most likely to shift from search to AI interactions over time
The queries that work best for AI prompt conversion are long-tail and question-based — typically five or more words, often starting with "what," "how," "which," "best," or "why." Short, fragmented queries like "CRM software" or "email marketing" are less useful as AI prompts because they don't reflect how people interact with ChatGPT.
## Connecting Google Search Console
Navigate to **Prompt Research** in the left sidebar and select **Search Console**. You will see the connection panel if you have not yet linked a GSC property.
Click **Connect Google Search Console**. You will be redirected to a Google OAuth screen. Sign in with the Google account that has access to the Search Console property for your site.
After authentication, GenRank displays the Search Console properties associated with your Google account. Select the property that matches the site you are tracking. If your site has multiple properties (for example, separate `www` and non-`www` versions), choose the one that captures the most data — typically the verified domain property.
Choose the date range of query data to import. A 90-day window gives you a broad view of query patterns. A shorter window (28 days) reflects more recent behavior if your site or category has changed recently.
Click **Import**. GenRank fetches your GSC query data and filters it for conversational, multi-word queries. The results appear as a list of prompt candidates ranked by click volume and query length.
## How GenRank identifies AI-ready queries
Not every query in your GSC data is a useful AI prompt. GenRank applies a filter to surface the ones most likely to map to genuine ChatGPT behavior.
Queries are surfaced as high-priority candidates when they:
* Contain five or more words
* Use question phrasing ("how to," "what is," "which is best," "can I use")
* Describe a specific use case, audience, or context rather than a generic category
* Already receive meaningful click volume, indicating real user intent behind the query
Queries that are typically filtered out include:
* Brand name + single keyword combinations (navigational intent)
* Queries with fewer than three words
* Queries that appear to be typos or encoding artifacts
Even if a query has low click volume, it may still be worth tracking as an AI prompt if the phrasing is highly conversational. Low-volume long-tail queries in traditional search are often the exact queries users bring to ChatGPT instead.
## Converting queries into tracked prompts
Once your GSC data is imported, the interface displays a filterable list of query candidates.
Browse the imported queries. Each entry shows the original query text, its click volume from GSC, and a suggested intent category assigned by GenRank. You can sort by volume, query length, or intent category.
GSC queries are raw search strings — they may need slight rewording to function well as ChatGPT prompts. For example, "best CRM small sales team" might become "what is the best CRM for a small sales team." Click any query to edit it before adding it to your library.
Check the queries you want to track and click **Add to Prompt Manager**. They are added to your library with intent categories pre-assigned, ready for you to activate.
Importing a large volume of queries from GSC does not automatically activate them as tracked prompts. Prompts only count against your slot limit once they are activated in Prompt Manager. Review and activate selectively to make the most of your available slots.
## Keeping your data fresh
Your GSC query landscape changes over time as your content evolves, new pages are indexed, and search behavior shifts. Revisit the Search Console integration periodically — monthly or quarterly — to surface new conversational queries that have emerged since your last import.
This usually reflects the nature of your site's current content and how it is indexed. Sites with short, keyword-dense pages tend to attract fragmented query traffic. Sites with detailed guides, comparison pages, and FAQ-structured content attract longer, more conversational queries. If your GSC data is sparse, Page Scan is a better starting point for building your initial prompt library.
Each GenRank project is connected to one GSC property at a time. If you manage multiple brands or domains, set up separate GenRank projects for each one.
No. GenRank requests read-only access to your Search Console query data specifically. It does not access Google Analytics, Google Ads, or any other Google property.
GenRank checks for duplicates during the import process and flags any GSC queries that match prompts already in your library. You can choose to skip these or import them anyway if you want to track slight variations separately.
# Get started with GenRank
Source: https://docs.genrank.io/quickstart
Set up your GenRank account, create a project, add competitors and prompts, and see your brand's ChatGPT visibility in under ten minutes.
This guide walks you through everything you need to do to get your first GenRank project running. From creating your account to seeing your brand's AI visibility data, the full setup takes under ten minutes.
Go to [app.genrank.io/onboarding](https://app.genrank.io/onboarding) and sign up. You can start on the Free plan with no credit card required — it includes 10 prompt slots and daily tracking so you can validate the platform before upgrading.
If you are setting up GenRank for a client or managing multiple brands, consider starting on the Pro plan, which supports up to 3 projects and 10 tracked competitors.
After signing in, you will be prompted to create your first project. A project is a tracking configuration tied to a single brand and domain.
Fill in the following fields:
* **Brand name** — The name of your brand exactly as you want it recognized (e.g., `Acme` not `Acme Inc.` or `ACME`). This is used to detect mentions in ChatGPT responses, so consistency matters.
* **Domain** — Your primary website domain (e.g., `acme.com`). GenRank uses this to track source citations and content retrieval alignment.
* **Geographic target** — The market or region you want to track visibility in. AI responses can vary by locale, so this scopes your data to the right audience.
You can edit these settings later from the project settings page. If your brand name has common abbreviations or alternate spellings, you can add them after the initial setup.
On the Essential plan and above, you can add competitors to benchmark against. GenRank will detect competitor mentions in the same ChatGPT responses it captures for your prompts.
Enter each competitor's brand name. You do not need to add their domain unless you also want to track their citation sources.
A few guidelines for choosing competitors:
* Include direct competitors that appear in the same category queries as your brand.
* Add aspirational competitors — brands you want to displace in AI recommendations.
* Avoid adding too many at first. Start with 3–5 and expand as you learn which ones dominate your prompt landscape.
The Free plan does not include competitor tracking. You will need the Essential plan (3 competitors) or Pro plan (10 competitors) to see competitor data.
Prompts are the questions GenRank sends to ChatGPT on your behalf. The prompts you track determine all the visibility data you receive, so choosing the right ones matters.
You have three ways to add prompts:
**Option A: Add prompts manually**
Type in prompts that your customers are likely to ask ChatGPT. Write them as natural, conversational questions rather than keyword fragments.
```text theme={null}
Good: "What is the best project management software for a remote startup?"
Avoid: "project management software startup"
```
Tag each prompt by intent to keep your data organized:
* **Discovery** — "What tools exist for X?"
* **Comparison** — "What is better, X or Y?"
* **Transactional** — "Which X should I buy?"
* **Branded** — "Tell me about \[your brand name]"
**Option B: Use Page Scan**
Enter a URL from your site — a product page, service page, or landing page — and GenRank will generate a list of prompts a user might ask ChatGPT about that content. Review the suggestions and add the ones that match real purchase intent in your category.
**Option C: Import from Google Search Console**
If you have Google Search Console connected to your domain, GenRank can pull long-form, question-based queries from your search data and surface them as ready-to-add prompts. These are queries users already type into search engines — many of which are migrating to AI tools.
Start with 5–10 well-chosen prompts rather than filling all your prompt slots at once. A focused set of high-intent prompts gives you cleaner data and faster learning. You can always add more once you see the initial results.
After you save your prompts, GenRank queues them for its next daily run. The first results are typically available within 24 hours.
When the data arrives, your dashboard will show:
* **Brand Visibility** — The percentage of your tracked prompts where your brand was mentioned, and how that breaks down prompt by prompt.
* **Share of voice** — How your mention frequency compares to each tracked competitor.
* **Response samples** — The full text of each ChatGPT response, with your brand mentions highlighted.
* **Sources** — Which external domains were cited in responses across your prompt set.
**Reading your first results**
A few things to keep in mind when interpreting early data:
* A single day's data is a starting point, not a trend. Visibility scores become more meaningful after 7–14 days of history.
* If your brand is absent from a prompt's response, that is a gap — not a failure. Prompt gaps are one of the most actionable signals in GenRank.
* Competitor mentions in prompts where you are absent are the highest-priority prompts to focus on for optimization.
GenRank captures real ChatGPT responses — not estimates or modeled predictions. Every mention and citation in your dashboard comes from an actual response that was recorded and stored.
## What to do next
Once you have your first week of data, you have a clear picture of where you stand in AI-generated answers. From here, the most productive next steps are:
Dive deeper into brand visibility, competitor gaps, and brand perception data.
Use Content Retrieval and Entity Clarity to understand why you appear — or don't.
Expand your prompt library with Page Scan or Search Console data.
Learn the key terms and ideas behind how GenRank measures AI visibility.
# Perception
Source: https://docs.genrank.io/response-tracking/brand-perception
See recurring attributes ChatGPT associates with your brand, compare positioning against competitors, and align content with the qualities you want to own.
Language models don't just name brands — they characterize them. When ChatGPT mentions your brand in a response, it typically frames it in terms of specific qualities: "easy to use," "enterprise-grade," "affordable," "reliable," "technical." These attributes aren't random. They reflect patterns in the training data and live web content that the model draws on, and they repeat consistently across different prompts and response runs. Brand Perception extracts and quantifies those patterns so you can see how AI actually talks about your brand at scale.
## What perception attributes are
A perception attribute is any recurring descriptive quality that ChatGPT associates with your brand name in captured responses. GenRank identifies these by analyzing the language used in close proximity to your brand mentions across all captured responses and surfacing the qualities that appear with statistically meaningful frequency.
Examples of perception attributes include:
* **Functional qualities**: "easy to set up," "integrates with existing tools," "API-first," "no-code"
* **Value positioning**: "affordable," "enterprise pricing," "cost-effective for small teams"
* **Trust signals**: "well-documented," "reliable uptime," "strong support"
* **Audience framing**: "popular with developers," "used by large enterprises," "suited for beginners"
* **Comparative positioning**: "more flexible than alternatives," "simpler than \[competitor]"
These are the terms that shape how a user's first impression of your brand forms when they encounter it in a ChatGPT answer — before they've visited your site or read any of your own marketing.
## How perception data is surfaced
The Brand Perception view shows your most frequently occurring attributes ranked by appearance frequency across your full prompt set. Each attribute entry includes:
* The attribute phrase as extracted from responses
* How many captured responses it appears in
* Which prompts it's most associated with
* How the frequency has trended over recent weeks
You can also compare your attribute profile against competitors. This side-by-side view shows which qualities appear for your brand, which appear for specific competitors, and which are contested — appearing for both.
## Using perception data
Understanding your current perception profile is the first step. The more valuable question is whether that profile matches what you want it to be.
Map your perceived attributes against your intended positioning. If your target positioning emphasizes enterprise reliability but your most frequent ChatGPT attributes are "affordable" and "suited for small teams," there's a gap between how you present yourself and how the broader information environment — which the model reflects — frames you. That gap points directly at content and PR priorities: you need more authoritative coverage that frames you in enterprise contexts, from the kinds of sources ChatGPT treats as credible.
### Aligning content with desired attributes
Perception attributes in ChatGPT responses are downstream of the information environment your brand exists in — your own content, third-party reviews, editorial coverage, community discussions, and citations across the web. Shifting your perception profile means changing the inputs the model uses.
Compare your current top attributes in GenRank against the attributes you want to be associated with. Note which desired qualities are absent or underrepresented, and which undesired qualities appear more than you'd like.
Use the Sources view to identify which domains are being cited in responses about your brand. Those sites are contributing most directly to how the model characterizes you. Read the content on those pages to understand what language they use about you.
Create content — on your own site and through earned coverage — that explicitly frames your brand in the attributes you want to own. Case studies, technical documentation, comparison content, and third-party reviews that use the right language consistently are the inputs the model will draw on.
Monitor your perception attribute frequencies in GenRank over the following weeks. Attribute shifts are typically slower than mention rate changes, but they are measurable. A rising frequency for a target attribute confirms that the information environment is starting to reflect your updated positioning.
### Identifying misalignments
Perception misalignments — attributes that don't match your intended brand position — can surface in a few different ways:
**Outdated positioning**: If your brand went through a repositioning, a pricing change, or a product pivot, the model may still reflect the older framing because the underlying content landscape hasn't fully updated yet.
**Category assumption**: The model may apply category-level attributes to your brand because of who you're typically compared to, even if those attributes don't accurately describe you.
**Competitive contamination**: If a competitor is closely associated with a particular attribute, some of that association may bleed into how the model frames the broader category — and by extension, you.
Identifying these misalignments gives you a concrete starting point for content corrections.
## Comparing perception across competitors
The competitor perception view shows how your attribute profile compares to each tracked competitor side by side. This reveals:
* Attributes you own exclusively (a potential differentiation signal)
* Attributes your competitor owns that you don't appear in at all (a positioning gap or a deliberate differentiation)
* Attributes both brands share (contested ground, where differentiation is less clear in AI responses)
Perception data is derived from the same captured responses as all other Response Tracking metrics. Attributes are extracted from the full response text associated with your tracked prompts — they reflect what ChatGPT says about your brand in the context of the questions your target audience is actually asking.
# Visibility
Source: https://docs.genrank.io/response-tracking/brand-visibility
Measure how often your brand is mentioned and cited in ChatGPT responses, track daily visibility trends, and review the exact responses behind the numbers.
Brand Visibility shows you the raw facts of your presence in ChatGPT: how many of your tracked prompts produce a response that mentions your brand, how many cite your domain as a source, and how that performance changes day over day. All figures come from captured ChatGPT responses — there is no modeling or estimation involved. You can trace every metric back to a specific response and read the exact text that produced it.
## Core metrics
### Mention rate
Mention rate is the percentage of your tracked prompts for which ChatGPT's response includes your brand name. A mention rate of 60% means your brand appears in 6 out of every 10 prompt responses GenRank captured.
Mention rate is the most direct indicator of AI discoverability. If ChatGPT does not name you in a response, a user following that conversation has no signal that you exist — regardless of how strong your traditional search presence is.
### Citation rate
Citation rate is the percentage of tracked prompts where ChatGPT cites your domain as an external source. This happens when ChatGPT uses web search to support its answer and includes your site in the citations table.
Citation rate measures a different kind of presence than mention rate. Your brand can be mentioned without a citation (the model draws on training data) and cited without a prominent mention (the source appears in the table but not in the main answer text). Tracking both separately tells you whether your domain is earning authority as a reference point, not just name recognition.
### Visibility percent
Visibility percent is a composite measure of how frequently and prominently your brand appears across all captured responses. It weights mention rate and citation rate together and accounts for whether your brand appears once in passing or multiple times in a substantive role. A brand that is mentioned briefly in one sentence and a brand that is discussed across several paragraphs both count as "mentioned" — visibility percent captures the difference.
## The dashboard
The Brand Visibility dashboard gives you three levels of detail.
**Daily trend chart**: The top of the dashboard shows your mention rate, citation rate, and visibility percent plotted over time. Each data point represents one day's full prompt run. Use this view to spot upward and downward movements and connect them to things you've done — published content, earned press coverage, launched a PR campaign, or updated your website.
**Per-prompt breakdown**: Below the trend chart, each tracked prompt is listed with its individual mention rate, citation rate, and latest visibility score. Sorting by mention rate shows you which prompts are already working. Sorting by citation rate shows where you're earning source authority. Prompts with low scores on both dimensions are your clearest optimization targets.
**Full response review**: Click any prompt row to open the response detail panel. You'll see the full ChatGPT response text, your brand name highlighted wherever it appears, the citations table if one was generated, and the perception attributes extracted from that response. This is the ground truth behind every metric — the actual output ChatGPT produced for that prompt on that day.
Use the daily trend view to measure the impact of specific actions. When you publish a new article, earn a mention in a major publication, or update your product page, note the date and watch whether your mention rate or citation rate shifts in the following days. Because GenRank refreshes data daily, you'll typically see the effect within one to two days if it's going to register. This turns Brand Visibility into a feedback loop for your content and PR work, not just a passive reporting tool.
## What the data is based on
Every figure in Brand Visibility is derived from responses that GenRank actually captured from ChatGPT. When GenRank runs your tracked prompts, it submits each one to ChatGPT, records the full response, and parses that text for mentions and citations. The stored response is always available for review — you are never looking at an estimate of what ChatGPT might say.
This matters because ChatGPT responses are not static. The model's outputs shift as its training data is updated, as web search results change, and as the broader information environment in your category evolves. Daily capture means you're measuring what the model actually says, not a historical approximation.
Mention rate and citation rate are calculated across **all active tracked prompts** in your project. If you add or remove prompts, the aggregate figures will change to reflect the new prompt set. Per-prompt history is preserved even if a prompt is later deactivated.
## Reading per-prompt results
ChatGPT's responses vary by prompt intent. Informational prompts ("what is...") often produce different brand distributions than comparison prompts ("which tool is best for...") or recommendation prompts ("what should I use to..."). Prompts with strong commercial intent in your category tend to have higher brand mention density overall. If your brand is absent from high-intent prompts, that's where visibility efforts have the most leverage.
Citations only appear when ChatGPT uses web search to support its answer. For prompts where the model answers from training data alone, no citations are generated regardless of whether your brand is mentioned. Citation rate measures a subset of responses — those where live web retrieval was triggered — so it will typically be lower than mention rate for most brands.
This typically means your brand is mentioned prominently and in substantive detail in the responses where it does appear, even though it doesn't appear in every prompt. ChatGPT is giving you significant coverage on the prompts where you show up — the opportunity is to increase the range of prompts on which you appear.
# Brands
Source: https://docs.genrank.io/response-tracking/competitors
Track competitor mentions across your full prompt set, compare share of voice, and find the exact prompts where rivals appear and your brand does not.
ChatGPT distributes attention across brands unevenly. Some names appear consistently across a wide range of prompts; others dominate specific topic clusters; others are largely absent. The Competitors view gives you a continuous, daily picture of how that distribution looks across your tracked prompt set — who is gaining ground, who is fading, and precisely where each competitor is capturing the visibility you're not.
## What competitor tracking covers
For every tracked prompt, GenRank records all brand names that appear in the ChatGPT response — not just yours. This produces a complete map of brand co-occurrence across your prompt set, updated daily. From that data, GenRank calculates:
* **Competitor mention count**: How many of your tracked prompts mention each competitor.
* **Share of voice**: Each brand's mentions as a percentage of all brand mentions across the full prompt set. This is the closest equivalent to market share within the AI answer layer.
* **Prompt ownership**: The prompts where a given competitor is mentioned and you are not — their territory, in practical terms.
* **Prompt overlap**: The prompts where both your brand and a competitor appear in the same response.
* **Prompt gaps**: The subset of prompt ownership that represents direct competitive displacement — prompts where a competitor appears but you don't, meaning a user following that conversation sees them and not you.
The number of competitors you can track depends on your plan: **Essential** supports up to 3 competitors, **Pro** supports up to 10, and **Scale** plans offer custom limits. Competitors are configured when you set up your project and can be updated from your project settings.
## Reading the competitor dashboard
The competitor dashboard has two main views.
**Leaderboard view**: Shows all tracked competitors ranked by mention count or share of voice, with a trend indicator showing whether each is rising or falling compared to the previous period. Use this view to identify who is dominating the conversation in your category right now and whether the competitive landscape is shifting.
**Per-competitor detail**: Click any competitor to see their prompt-by-prompt performance — where they appear, how often they're cited, what share of voice they hold on individual prompts, and where your visibility overlaps with or diverges from theirs.
## Using prompt gaps
Prompt gaps are the most actionable output of competitor tracking. A prompt gap is any prompt where a competitor is mentioned and your brand is not. It represents a user intent your category is actively serving — but your brand is being excluded from the answer.
Think of prompt gaps as a prioritized content brief. Each gap tells you exactly which question ChatGPT is already answering with a competitor's name attached. Those are the prompts where targeted content improvements, stronger entity signals, or earned media in that topic area are most likely to shift your inclusion rate.
### Example: using prompt gaps in practice
Suppose you run a project management SaaS. Your tracked prompts include a mix of broad category queries and specific use-case questions. Here's how prompt gap analysis might look in practice:
Your prompt gap report shows that a competitor appears in 18 prompts where you don't. Drilling into those 18 prompts, you notice a cluster of 7 are variations of "best project management tool for remote teams" and "how to manage distributed teams with software." Your brand does not appear in any of them. The competitor appearing in those gaps has several recent case studies and blog posts specifically about remote team workflows.
ChatGPT is treating your competitor as the authority on remote team use cases. Your content may not signal sufficient relevance to that specific sub-topic, or you may not have enough source coverage (articles, reviews, mentions) that frames you in that context. The model's answer reflects what the broader information environment says — and right now, it says your competitor owns that territory.
Publish content that directly addresses remote team workflows and positions your product in that context. Earn coverage in publications that cover distributed work. Update your product page to make remote-team capabilities explicit. When GenRank's next data run captures those prompts, you'll be able to measure whether the changes are moving your inclusion rate on that specific cluster of gaps.
## Tracking competitive movement over time
The competitor view is designed for continuous monitoring, not one-time analysis. Share of voice figures update daily, so you can see when a competitor starts appearing in more prompts — often weeks before that translates into detectable business impact. Early signals include a rising mention count, an expanding prompt footprint, or new citations from domains you haven't seen them associated with before.
Similarly, you'll see when competitors lose ground. A brand that was consistently appearing in 40% of your prompts and drops to 20% over a two-week period has experienced a meaningful shift. Understanding why — a change in their content, a loss of key third-party citations, a shift in how the model categorizes them — is the kind of competitive intelligence that lets you act proactively rather than reactively.
Any instance of a competitor's brand name appearing in a captured ChatGPT response to one of your tracked prompts. GenRank uses the brand names you specify during project setup and matches them against the full response text. Each unique response containing the brand counts as one mention for that prompt run.
Yes. You can update your competitor list from your project settings. Changes take effect on the next daily data run. Historical data for newly added competitors will not be backfilled — tracking starts from the date you add them.
GenRank tracks only the brands you've explicitly added as competitors. If you notice a brand name appearing in your response review panel that you haven't tracked, you can add them to your competitor list to begin measuring them systematically.
# Overview
Source: https://docs.genrank.io/response-tracking/overview
Measure how your brand appears in real ChatGPT responses — mentions, citations, competitor presence, and perception — updated daily across tracked prompts.
Response Tracking is GenRank's core measurement module. It captures real ChatGPT responses to your tracked prompts every day and structures the output into four distinct lenses: how often your brand is mentioned and cited, how competitors compare, how AI characterizes your brand, and which external sources shape the answers in your category. Every data point comes from actual captured LLM responses — not modeled estimates or projections.
## Why AI response tracking matters
When someone asks ChatGPT for a product recommendation, a service comparison, or expert advice, the answer they receive shapes their perception and buying intent. Search engines return ten links and let users decide. ChatGPT selects a small set of brands and presents them as authoritative. If your brand is absent from that selection, you lose visibility in a channel that is growing faster than traditional search.
Response Tracking makes that selection process visible. You see which prompts mention you, which cite your domain, which favor competitors, and what language the model uses to describe you. That visibility is the foundation for any meaningful GEO strategy.
## How data is collected
GenRank runs each of your tracked prompts in ChatGPT on a daily schedule. The full response is captured, stored, and parsed for brand mentions, citation links, perception attributes, and web search queries. Nothing is inferred from training data or estimated from proxy signals — every metric is derived directly from the response text.
GenRank submits each tracked prompt to ChatGPT and captures the complete response output, including any web search queries the model performs before answering.
Each response is analyzed for brand mentions, domain citations, recurring attribute language, and competitor co-occurrence. The raw response text is also stored so you can read it in full.
Daily results are combined into time-series metrics — mention rate, citation rate, share of voice, and more — so you can see changes over time, not just a one-time snapshot.
GenRank flags prompt gaps where competitors appear but you do not, and highlights when your visibility scores move significantly in either direction.
All metrics are refreshed **daily** for every tracked prompt. Changes from content updates, PR coverage, or SEO work typically become visible within one to two days of GenRank's next data run.
## Key metrics
The percentage of your tracked prompts where your brand name appears in the ChatGPT response.
The percentage of tracked prompts where your domain is listed as a cited source in the response.
A combined measure of how frequently and prominently your brand is mentioned across responses, not just whether it appears.
Your brand's mentions as a proportion of all brand mentions across your full prompt set — the AI equivalent of market share in search results.
## What's included in Response Tracking
Track your mention rate, citation rate, and visibility percentage across all prompts with daily trend data and per-prompt breakdowns.
See how competitors are distributed across your prompt set, identify who is gaining ground, and surface prompt gaps where they appear and you don't.
Discover the recurring attributes ChatGPT associates with your brand and compare your positioning against competitors.
Track which external domains ChatGPT cites in responses to your prompts, and see how your own domain performs within that ecosystem.
## How this differs from SEO tracking
Traditional SEO tools track your position in a ranked list of ten links. Response Tracking measures something structurally different: inclusion in an AI-generated answer that names only a few brands. There are no positions to rank — only presence or absence, and the frequency and framing of that presence. Share of voice, prompt gaps, and perception attributes are metrics that have no direct equivalent in SEO, because the format of AI answers creates a different competitive dynamic.
GenRank currently tracks responses from **ChatGPT**, which holds over 69% market share among AI assistant users. Additional LLM coverage is on the roadmap.
# Sources
Source: https://docs.genrank.io/response-tracking/sources
See which external domains ChatGPT cites for your tracked prompts, track your domain's citation rate, and focus content and PR efforts where they matter most.
When ChatGPT uses web search to support an answer, it doesn't just generate text — it pulls from specific external pages and lists them as citations. Those cited domains are the information sources the model treats as credible and relevant for that query. The Sources view tracks exactly which domains are cited in responses to your tracked prompts, how often each one appears, and which brands they're associated with. This gives you a direct view into the third-party information ecosystem that shapes AI answers in your category.
## What "sources" means in this context
A source is any external domain that ChatGPT cites in a response to one of your tracked prompts when the model performs a web search. Citations appear in the response as a table or inline links, referencing the pages the model consulted before generating its answer.
Not all ChatGPT responses include citations. The model only performs web search for certain prompt types — typically those requiring current information, product comparisons, or recent events. When it does, the cited pages represent the live web content that directly influenced the response text. GenRank logs those citations on every prompt run that produces them.
## Why source tracking matters
The domains that get cited are the domains that have a seat at the table when ChatGPT forms its answer. A review site that consistently appears in citations for comparison prompts in your category is shaping what the model says about those products — including yours. A media publication that covers your industry and frequently gets cited has more influence over your AI visibility than a site that writes about you but never gets cited.
Understanding the citation ecosystem lets you move beyond your own domain and understand the full information environment you're operating in. You can see:
* Which third-party sites carry the most weight for prompts in your category
* Whether the sites that cover you positively are actually being cited
* Which sources your competitors are drawing citation support from
* Where your own domain stands relative to the rest of the ecosystem
## How your domain performs as a source
The Sources view shows your own domain's citation metrics alongside all other cited domains. Key figures include:
**Your citation count**: How many times your domain appears in citations across all captured responses. This is the raw frequency of your domain being treated as a credible source.
**Your citation rate**: The percentage of responses that generated citations where your domain was included. This normalizes against the total number of cite-eligible responses rather than all responses.
**Prompts you're cited on**: The specific tracked prompts where your domain appears in citations. This tells you which question contexts ChatGPT currently treats your site as a relevant source for.
**Prompts you're missing from**: Tracked prompts that generate citations but don't include your domain, even when your brand is mentioned in the response text. This gap — mentioned but not cited — means ChatGPT is drawing on your brand's reputation from training data but not treating your own site as an authoritative reference.
## Using source data to guide content and PR strategy
Sort the sources list by citation frequency across your tracked prompts. The domains at the top are the ones ChatGPT most consistently turns to when forming answers in your space. These are the publications, review sites, directories, and communities that carry disproportionate weight.
Check whether your brand is mentioned — and mentioned accurately and favorably — on the highest-citation domains. If a site appears in 30% of your tracked prompt citations but doesn't mention your brand, or covers you only briefly, that's a concrete PR and content gap.
Compare the prompts where your domain is cited against those where it isn't, especially for prompts where you're already mentioned in the response text. Being mentioned but not cited means you're present in the model's general knowledge but not earning source authority on that topic.
Focus content investment on the topics and prompt clusters where the citation ecosystem is active — where other domains are regularly getting cited — but your site isn't among them. Those are the areas where publishing strong, citable content has the clearest path to improving your citation rate.
After publishing new content or earning coverage on a high-influence domain, monitor your citation rate on the relevant prompts. Because GenRank tracks citations daily, you can see when a new page starts getting picked up as a source.
Being mentioned and being cited are two different things with different leverage points. Increasing your mention rate is primarily a question of brand presence in the information environment — how well-known you are in your category. Increasing your citation rate is a question of content authority — whether the specific pages on your domain are treated as credible references for specific query types. You can improve citation rate even on prompts where your mention rate is already high by publishing content that directly addresses those prompt topics in a format ChatGPT finds citable.
## Reading the web search queries
Alongside citations, GenRank also logs the web search queries ChatGPT executes before generating a response. These queries reveal how the model decomposes a user's prompt into specific information-seeking steps.
For example, a user prompt like "what's the best CRM for a small consulting firm?" might cause ChatGPT to run searches like "CRM software for consulting firms 2025," "best CRM small business comparison," and "CRM pricing small team." Each of those queries represents a distinct information angle the model is trying to cover — and each is an opportunity for your content to appear in those search results and subsequently get cited.
Web search queries are logged only for prompt runs where ChatGPT performed a web search. Not every prompt triggers a search. The proportion of search-enabled responses varies by prompt type and changes over time as ChatGPT's behavior evolves.
## Connecting sources to your broader strategy
The Sources view ties together Response Tracking and your upstream content and PR work. Mentions in AI responses come from somewhere — from training data, from live web search, from cited third-party coverage. Source tracking shows you the live-web layer of that equation: which pages are being retrieved, which domains are being trusted, and where your site fits in that ecosystem.
Not necessarily, but citation rate is a strong signal of source authority. A high citation rate means ChatGPT is treating your domain as a credible reference for those prompt topics. Brands with high citation rates tend to also have higher visibility percent scores, because cited content directly shapes response text. Citation rate and mention rate together give a more complete picture than either metric alone.
Yes. The per-prompt response detail shows the full citation table from that response run, including the specific URLs that were cited. This lets you see exactly which pages are earning citations and which topics they're associated with.
This is one of the most valuable signals the Sources view surfaces. A frequently cited domain that mentions competitors but not your brand is a clear PR target. Getting covered on that site — especially in content that explicitly compares or recommends products in your category — is one of the most direct ways to improve your citation rate on the prompts where that domain appears.