URL: /capabilities/check-ai-answers · Id: `page:/capabilities/check-ai-answers`
# Check AI answers
> What the answer engines say when a buyer asks for what you make: the brands they name and the pages they cite.
What does ChatGPT say about my category? Your agent asks each answer engine the questions your buyers ask, then lists the brands it names and the pages it cites.
```text title="Check AI answers · Ask"
Ask ChatGPT and Perplexity what a landlord with 50 units should use to track rent.
Tell me who they name, which pages they cite, and check again every week.
```
```ts title="Check AI answers · SDK"
import { $ } from 'startup-builder'
const chatgpt = await $.Keyword.measure({ keyword: 'keywords/keyword_21', venue: 'chatgpt' }) // “best way to track rent for 50 units?” on ChatGPT
const perplexity = await $.Keyword.measure({ keyword: 'keywords/keyword_21', venue: 'perplexity' }) // the same question on Perplexity
await $.Monitor.create({ kind: 'outside', target: 'keywords/keyword_21', every: '7d' }) // an Answer watch on every approved question, weekly
return { chatgpt, perplexity }
```
```json title="Check AI answers · Response · sample"
{
"$id": "https://api.sb/leaseline.app/keywords/keyword_21",
"$type": "Keyword",
"question": "What is the best way to track rent for 50 units?",
"answers": [
{
"venue": "chatgpt",
"observedAt": "2026-10-25",
"named": [
"RentRoll",
"LedgerNest",
"a spreadsheet template"
],
"cited": [
"rentroll.example/pricing",
"reddit.com/r/Landlord",
"g2.com/categories/property-management"
],
"searchesRan": [
"rent tracking software small landlord",
"rentroll vs ledgernest"
],
"you": "not named"
},
{
"venue": "perplexity",
"observedAt": "2026-10-25",
"named": [
"LedgerNest",
"RentRoll",
"KeyPlace"
],
"cited": [
"ledgernest.example/blog/rent-ledger",
"reddit.com/r/Landlord",
"keyplace.example/features"
],
"you": "not named"
}
],
"cost": {
"usd": 0.0099
}
}
```
| First run | Status | Runs by | Connect once |
| --- | --- | --- | --- |
| 20 questions on 4 answer engines: about $0.30 | Coming | API · browser agent | Nothing to start. Your Bing sign-in, once, adds the Copilot citations report. |
## What you get
| Field | What it says |
| --- | --- |
| Question | A question a buyer asks, phrased the way your buyers phrase it in search and in posts. |
| Engine | ChatGPT, Gemini, Perplexity, Copilot, Claude or Google AI Mode, each read on its own. |
| Answer | The answer as a buyer sees it, signed out, from your Market, dated. |
| Brands named | Every brand the answer names, in order, with you marked when you appear. |
| Pages cited | The pages the answer cites, which become the places to earn a mention or a link. |
| Searches it ran | The searches the engine ran to build the answer, which become terms in Research demand. |
| Answer share | The share of answers that name you, per engine, read each week. |
| Cost | The most the check can cost, shown before it runs, and what it cost after. |
## How it runs
| When | What happens | Runs by | Your part | Takes |
| --- | --- | --- | --- | --- |
| Minute 1 | Build the question set from the questions people also ask and the problem posts | API | approve 20 questions | 5 minutes of yours |
| Minute 2 | Ask each question on each answer engine, signed out, from your Market ([ChatGPT answer](#chatgpt)) | API | nothing | under a minute an answer |
| Minute 3 | Tally the brands named and the pages cited across every answer ([Most-cited sites](#most-cited-sites)) | API | nothing | the same minute |
| Day 1 | Hand the cited pages to Earn first links as places to earn a mention | API | nothing | the same day |
| Weekly | Ask the approved questions again and flag a change in who is named or cited ([Answer watch](#answer-watch)) | API | nothing | the same day; a new page of yours can show once it is indexed |
| Weekly | Read the Copilot citations report on Bing, signed in as you ([Copilot citations](#copilot-citations)) | browser agent | one sign-in, once | once Bing has data on your site |
## Jobs inside it
| Job | What it returns | Price | Status | Runs by |
| --- | --- | --- | --- | --- |
| ChatGPT answer | The answer, the brands named, the pages cited and the searches it ran | $0.004 a question | Coming | API |
| Gemini answer | The answer and its sources | $0.0024 a question | Coming | API |
| Perplexity answer | The answer and its citations | $0.0059 a question | Coming | API |
| Copilot answer | The answer, its sources and the places it names | $0.0028 a question | Coming | API |
| Claude answer | The answer and the searches it ran | $0.0245 a question | Coming | API |
| Google AI Mode answer | The AI answer above the results and the pages it cites | $0.0024 a question | Coming | API |
| Most-cited sites | The sites answer engines cite most for your topic: where a mention counts | about $0.10 a topic | Coming | API |
| Answer watch | Your approved questions asked again each week, with a change in who is named flagged | the answer prices, once a week | Coming | API |
| Copilot citations | Bing’s report of your cited pages and the queries behind them | free | Coming | browser agent |
## Where and how it runs
| Where it works | How it runs | Connect once |
| --- | --- | --- |
| ChatGPT, Gemini, Perplexity, Claude, Google AI Overviews, Bing | API · browser agent | Nothing to start. Your Bing sign-in, once, adds the Copilot citations report. |
Your agent acts as you on every Venue you authorize. Where a Venue has no API, a browser agent signs in as you and does the steps.
## Plays that use it
| Play | Strategy that runs it |
| --- | --- |
| [Answer engine optimization](/plays/aeo) | [Passive](/strategy/passive) |
## Reference
| Tool | Runs today |
| --- | --- |
| | Coming |
| | Coming |