AI Brand Visibility: How to Track Your Brand in AI Overviews, ChatGPT and Gemini

AI brand visibility is how often AI answer engines such as Google AI Overviews, Google AI Mode, ChatGPT, Gemini and Copilot name your brand when buyers ask about your category, in what position, in what tone, and whether they cite your website. You measure AI brand visibility by sending the same buyer prompts to each engine on a schedule and recording mentions, rank, citations and share of voice against your competitors.

This guide shows a real check of HubSpot on 23 October 2026 (the same prompt on Gemini and Copilot gave opposite answers), the metrics worth tracking, a weekly schedule that writes to Google Sheets and Slack, Python code, and what each check costs.

Why AI answers need their own tracking

A classic ranking report tells you where your page sits among ten links. An AI answer is different: it names a handful of brands, puts one first, and cites a few pages that may not be yours. Google describes how both AI Overviews and AI Mode "may use a 'query fan-out' technique — issuing multiple related searches across subtopics and data sources — to develop a response" (Google Search Central). The page that shapes the answer can be a review site ranking for a sub-question you never tracked.

Google also notes that AI Overviews "often don't trigger", because they appear only where the system finds them additive. So there are two questions to measure: does an AI answer appear for your keyword, and when it does, are you in it?

The metrics to track

The AI Brand Visibility Tracker returns the first seven for every prompt and engine, plus a 0 to 100 visibility score (50 for a mention, up to 30 for being named early, 20 for a cited domain). The Google AI Overviews Scraper covers the last one.

A real AI brand visibility check: HubSpot on Gemini and Copilot

On 23 October 2026 we asked Gemini and Copilot "best CRM software for small business" from the US, tracking HubSpot against Salesforce, Pipedrive and Zoho CRM. Both checks were answered in 15 seconds; the run took 18.7 seconds and was charged $0.0817.

2026-10-23T03:05:41.686Z [scrapely.actor] INFO  Checking 1 prompts on Google Gemini, Microsoft Copilot for 'HubSpot' with 3 competitors (country US, language en, proxy DATACENTER).
2026-10-23T03:05:56.600Z [scrapely.actor] INFO  Done: 2 of 2 checks answered in 15 s. HubSpot was mentioned in 50% of answers (share of voice 12%).

One prompt, two engines, opposite outcomes. That's the whole case for tracking every engine separately:

• Being named first isn't the same as owning the answer. Gemini put HubSpot at #1, yet Zoho CRM had more than twice HubSpot's share of voice in the same answer (7 mentions against 3).
• Copilot never mentioned HubSpot. Its answer led with Zoho CRM, Monday.com and Bigin, then Freshsales. Two of those weren't in our competitor list. Read a sample of full answers, then add the brands the engines actually name.
• Neither engine cited hubspot.com. Both leaned on third-party roundups (Forbes, Capterra, Business.com, a Slack blog post). Those pages are your outreach list: the README calls it outreach research, and the summary rows rank the top cited domains per engine.

Google AI Mode told a different story again. In a run for our guide on scraping Google AI Overviews, the Google AI Mode Scraper asked "best crm for startups" on 23 October: HubSpot was named first (5 mentions), Pipedrive second, Salesforce not at all, and AI Mode cited hubspot.com as a source in its answer. It returned 20 sources, 10 of them cited, plus the follow-up questions AI Mode suggested.

For coverage, the same guide's AI Overviews run checked three US searches ("how to choose a heat pump", "what is python programming", "kubernetes pros and cons"). All three triggered an AI Overview with 7 to 10 cited sources, Reddit appeared among the sources of all three, and two of the three overviews streamed in after the page loaded, which the Actor waits for and records as live.

Which Actor for which engine

Start with the tracker for the brand view; it covers all five engines at $0.04 a check. Add the AI Overviews Scraper for keyword coverage at scale, and the dedicated ChatGPT or AI Mode scrapers when you want the full answer record, product picks or follow-up questions.

Set up weekly tracking with Google Sheets and Slack

1. Write the prompts. Collect 10 to 20 questions your buyers ask: category questions ("best CRM for small business"), comparisons ("HubSpot vs Pipedrive"), alternatives ("HubSpot alternatives") and problems ("how to track sales leads"). Use the language and country of each market.
2. Save the input. On the tracker's page, enter your brand, domains, prompts, competitors (with Name | domain | alias) and engines, then Save as new… in the Input menu. Gemini and Copilot are fast; ChatGPT, AI Mode and AI Overviews run in a real browser and add browser time.
3. Schedule it. In Automations, click New automation, set When to On a schedule with Weekly, and add the saved input under Run.
4. Compare with last week. Turn on the AI step in Compare with last run mode: "Which prompts gained or lost a HubSpot mention since last week, and on which engine?" It's billed per use, like the AI chat.
5. Send it. Append the rows to Google Sheets (that sheet becomes your history and chart source) and post the AI text to Slack. Under Keep, Only rows where… type equals check keeps the per-check rows. The automations guide has every option.

Set Max cost per run with headroom for browser engines. The tracker reserves budget before it starts each check and never starts one it can't finish within your cap: in our first attempt, ChatGPT plus AI Overviews under a $0.10 cap started no checks at all and was charged $0.0007.

Run the checks with the API

Create an API token under Settings → API & Integrations and export it as SCRAPELY_TOKEN. A Gemini and Copilot check finishes in under 30 seconds, so the synchronous endpoint returns the rows directly:

curl -X POST "https://api.scrape.ly/v2/actors/querystone~ai-brand-visibility-tracker/run-sync-get-dataset-items" \
  -H "Authorization: Bearer $SCRAPELY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"brandName": "HubSpot", "brandDomains": ["hubspot.com"], "prompts": ["best CRM software for small business"], "competitors": ["Salesforce | salesforce.com", "Pipedrive | pipedrive.com", "Zoho CRM | zoho.com | Zoho"], "engines": ["gemini", "copilot"], "countryCode": "us"}'

This Python script (pip install scrapely-client-python) runs three prompts on two engines, prints mention rate, first-place rate and citation rate per engine, lists the most cited sources, and appends every check to a history file:

import csv
import os
from collections import Counter, defaultdict
from datetime import date
from scrapely_client import ScrapelyClient

client = ScrapelyClient(os.environ["SCRAPELY_TOKEN"], api_url="https://api.scrape.ly")

run = client.actor("querystone/ai-brand-visibility-tracker").call(run_input={
    "brandName": "HubSpot",
    "brandDomains": ["hubspot.com"],
    "prompts": ["best CRM software for small business", "HubSpot alternatives", "easiest CRM for a 5-person team"],
    "competitors": ["Salesforce | salesforce.com", "Pipedrive | pipedrive.com", "Zoho CRM | zoho.com | Zoho",
                    "Monday.com | monday.com", "Freshsales | freshworks.com"],
    "engines": ["gemini", "copilot"],
    "countryCode": "us",
    "includeAnswerText": False,
})
checks = [r for r in client.dataset(run.default_dataset_id).iterate_items() if r.get("type") == "check"]

by_engine = defaultdict(list)
for check in checks:
    by_engine[check["engineName"]].append(check)
for engine, rows in by_engine.items():
    answered = [r for r in rows if r.get("hasAnswer")]
    mentioned = [r for r in answered if r["brandMentioned"]]
    first = [r for r in mentioned if r["brandRank"] == 1]
    cited = [r for r in answered if r["brandDomainCited"]]
    print(f"{engine}: mentioned in {len(mentioned)}/{len(answered)}, named first in {len(first)}, site cited in {len(cited)}")

domains = Counter(d for check in checks for d in check.get("citedDomains", []))
print("Most cited sources:", domains.most_common(5))

with open("ai-visibility-history.csv", "a", newline="") as f:
    writer = csv.writer(f)
    for c in checks:
        writer.writerow([date.today().isoformat(), c["engine"], c["prompt"], c["brandMentioned"],
                         c["brandRank"], c["brandDomainCited"], c["visibilityScore"]])

On our saved run, the summary step prints "Google Gemini: mentioned in 1/1, named first in 1, site cited in 0" and "Microsoft Copilot: mentioned in 0/1, named first in 0, site cited in 0". Three prompts on two engines are six checks, $0.24 in fees.

AI agents can run checks over MCP. Add https://api.scrape.ly/mcp to Cursor or Claude Code and ask: "Use scrapely to check whether Gemini and Copilot mention HubSpot for 'best CRM software for small business' and which sites they cite."

What it costs

• Our tracker run: 2 checks × $0.04 = $0.08 in fees; $0.0817 charged in total.
• The AI Mode run: 1 answer × $0.15; $0.1746 in total, the rest browser time.
• The AI Overviews run: 3 overviews × $0.001 = $0.003; $0.0259 in total, mostly browser time.

A worked example: 20 prompts on three engines every week is 60 checks, or $2.40 a week in fees and $9.60 for four weekly runs, before platform usage (browser engines add browser time). Adding AI Overview coverage for 200 keywords costs at most $0.20 a week in fees. The Free plan's $5 of monthly credit covers a first audit; the Starter plan at $19 a month includes $19 of credit.

What to watch out for

• Answers vary. The same prompt can produce different answers from one week to the next. Judge mention rates across many prompts and several weeks, not one answer.
• Market and language matter. Engines answer from the country you set and cite local sources. Track each market with prompts in its own language.
• Aliases count. Sub-brands and products change the numbers: Copilot's mentions of Bigin, which it described as "by Zoho", counted toward Zoho CRM through the Zoho alias. Set aliases on purpose for your brand and your competitors.
• Coverage before visibility. Many keywords show no AI Overview at all. Track hasAiOverview so a drop in mentions isn't really a drop in overviews.
• Browser engines are slower. Gemini and Copilot answered in 8 and 15 seconds; ChatGPT, AI Mode and AI Overviews take longer and add browser time. Schedule them weekly rather than hourly.

How to get mentioned more often in AI answers

Google says that "there are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary": a page must be indexed and eligible for a snippet, and SEO fundamentals apply. In practice, the tracking data tells you where to work. Cited sources show which roundups and review sites engines trust for your category; getting listed and described accurately there is the most direct lever. Prompts where competitors are named and you aren't show the content gaps on your own site. Our guide to scraping Google search results shows how to see who ranks for the sub-questions behind those answers.

Why Scrape.ly

• Five engines, one Actor. Gemini, Copilot, ChatGPT, Google AI Mode and AI Overviews in a single run, with competitors, citations and share of voice per engine.
• $0.04 a check, no subscription. Pay per answered check; failed checks and prompts an engine doesn't answer are free, and a maximum charge per run caps spend.
• Raw answers, not a black box. Full answer text, every source and a transparent sentiment score, exportable to JSON, CSV, Sheets or your warehouse.
• Automation built in. Weekly schedules, an AI compare step, Slack and Google Sheets deliveries, plus MCP for agents.

FAQ

What is AI brand visibility?

It measures how AI answer engines present your brand: whether they mention it for your category's questions, in which position, with what tone, and whether they cite your website. It's tracked per engine and per prompt, against your competitors.

How do I check if ChatGPT mentions my brand?

Add your prompts and brand to the AI Brand Visibility Tracker with the ChatGPT engine, or use the ChatGPT Search Scraper with your brands in Brands to track. Each answer comes back with mentions, order of first mention and whether your site is cited.

What is AI search visibility?

AI search visibility is the search-side view of the same idea: whether AI features in search engines, such as Google AI Overviews and AI Mode, appear for your keywords and cite your pages. Track overview coverage and citations alongside brand mentions.

Can I scrape Google AI Overviews?

Yes. The Google AI Overviews Scraper returns the overview text and every cited source for each keyword, and charges only when an overview appears. See how to scrape Google AI Overviews, or the Google AI Overviews API page for the request.

How often should I check my AI visibility?

Weekly suits most brands: answers vary, so trends over several weeks say more than daily snapshots. Run an extra check after a launch, a pricing change or a major content push.

Do I need special markup to appear in AI Overviews?

No. Google states there's no special schema.org structured data or AI-specific file needed; the page must be indexed and eligible for a snippet, and normal SEO best practices apply.