An estimated one in three purchase research sessions now starts with an AI assistant rather than Google. When your potential customer types "what's the best project management tool for startups?" into ChatGPT, your brand either shows up — or it doesn't.
Traditional SEO dashboards track your Google rankings. None of them track whether AI recommends you.
This post shows you how to check your AI visibility in under five minutes, and how to set up automatic weekly tracking so you're always in the loop.
Why AI recommendations are different from Google rankings
When someone searches Google, they get a list of ten blue links and decide for themselves. When someone asks ChatGPT or Perplexity, they get a single confident recommendation — and most of them follow it.
The mechanic is different too. Google ranks pages. AI assistants draw on everything they've been trained on and everything the web-search mode returns in real time. A brand that's well-documented across Reddit, blog posts, GitHub, and review sites gets recommended more often than one with a polished website but thin presence elsewhere.
There are two distinct things you can measure:
- Brand mentions — Does the AI name your brand in response to buyer-intent queries?
- Citation rank — Does the AI cite your website as a source in its footnotes?
Both matter. A brand mentioned but not cited is being recommended but not treated as authoritative. A site cited but not mentioned is being used as evidence for a competitor's recommendation. The goal is both.
Step 1: Run a manual spot check
Start with the simplest possible version: open ChatGPT and Perplexity and ask the questions your customers ask.
Choose 3–5 buyer-intent queries. Good examples:
- "What's the best [category] tool for [use case]?"
- "How do I [problem your product solves]?"
- "What do companies use for [your space]?"
For a project management tool, you might ask:
- "What project management software should a 20-person startup use?"
- "Best free project management tool for remote teams"
- "Alternatives to Jira for small teams"
Run each query on ChatGPT (with web search on), Perplexity, and Gemini. Note:
- Was your brand mentioned at all?
- How early was it mentioned? First recommendation, third, or buried in a list?
- Was your website listed in the citations/sources?
- Which competitors got mentioned instead?
This gives you a snapshot. The problem with doing it manually is it takes 20+ minutes per check, you can't easily track it over time, and you forget the results by next week.
Step 2: Automate it with an Apify actor
To track this systematically — across multiple queries, multiple engines, and week over week — I built an actor that does the heavy lifting: the AI Brand Visibility Tracker.
It runs real buyer-intent queries across ChatGPT, Perplexity, Gemini, and Claude, then returns structured data for each check:
{
"brand": "Asana",
"engine": "chatgpt",
"query": "best project management tool for small teams",
"brand_mentioned": true,
"mention_count": 2,
"position_score": 9,
"share_of_voice": 0.4,
"competitor_mentions": { "Trello": 2, "Monday.com": 1, "ClickUp": 0 },
"cited_domains": ["g2.com", "blog.asana.com"],
"response_snippet": "For small teams, Asana offers a generous free tier..."
}
The position_score is 1–10: 10 means your brand was named first, 1 means it was mentioned last, 0 means it wasn't mentioned at all. share_of_voice is your mentions divided by total brand mentions in the response — a direct measure of how dominant you are in that answer.
Setting it up takes about 2 minutes:
- Go to the actor page on Apify and click Try for free (no credit card needed for your first runs)
- Enter your brand name and up to 5 competitor names
- Paste in 3–5 queries your customers actually use
- Select which engines to check: ChatGPT, Perplexity, Gemini, Claude
- Click Start
The run takes a few minutes and outputs a dataset you can download as CSV or connect to Google Sheets.
Step 3: Set up weekly tracking
A one-time check tells you where you stand today. Weekly tracking tells you whether your content and PR efforts are actually moving the needle.
In the Apify console:
- Open Actors → Schedules → New schedule
- Point it at the AI Brand Visibility Tracker with your saved input
- Set
trackTrends: true(this is the default) - Choose a weekly cadence — I run it every Monday morning
Every subsequent run automatically compares to the previous one and adds trend deltas to the SUMMARY output:
{
"Asana": {
"mention_rate": {
"current": 0.67,
"previous": 0.5,
"delta": 0.17
},
"avg_position_score": {
"current": 7.5,
"previous": 6.0,
"delta": 1.5
}
}
}
A rising mention_rate means AI is recommending you more often. A rising position_score means you're being named earlier — more prominently — in the response. A rising share_of_voice means you're taking ground from competitors.
Step 4: Also check if your site is being cited
Brand mentions and citation links are different signals — and you want both.
The companion actor is the AI Citation Auditor. Instead of tracking brand names, it tracks whether your domain appears in the structured citation list that engines like Perplexity and ChatGPT return alongside their answers.
The input is domain names instead of brand names:
{
"domains": ["yourdomain.com", "competitor1.com", "competitor2.com"],
"queries": ["best project management tool for small teams"],
"engines": ["chatgpt", "perplexity", "gemini"]
}
It tells you:
- Whether your domain was cited at all
- What rank your citation appeared at (rank 1 = first source listed)
- Which specific URL was cited (so you know which content is earning citations)
- How that changes week over week
One useful property: multiple domains share the same API call. Auditing 5 domains against 3 queries × 3 engines costs the same as auditing 1 domain — the actor checks all domains against each response.
What actually improves your AI visibility?
Based on what I've seen from running these checks across various brands:
What helps:
- Being discussed across multiple sources. AI assistants synthesise from many inputs. A brand with Reddit threads, G2 reviews, blog posts, and GitHub repos mentioning it in context gets pulled into more responses than a brand with only a polished website.
- Being cited by authoritative domains. G2, Capterra, Trustpilot, and niche industry blogs that get cited by AI are worth being listed on.
- Clear, specific positioning. Brands that are easy to describe in a sentence ("the Notion alternative built for engineers") get recommended more precisely than brands with vague positioning.
- Being named in comparison content. "X vs Y" articles, "best tools for Z" lists — AI engines love this format because it's already structured as a recommendation.
What doesn't move the needle as much as you'd hope:
- Having a well-optimised website alone
- Meta descriptions and title tags (these are Google signals, not AI signals)
- Domain authority in the traditional SEO sense
Tracking this for clients
If you're an SEO or GEO agency, this is a straightforward addition to your monthly reporting. Run the actor with a unique trendKey per client:
{
"brands": [{ "name": "ClientBrand", "competitors": ["Competitor1", "Competitor2"] }],
"queries": ["..."],
"trendKey": "client-acme-july-2026"
}
Each client's trend series is stored separately. You get a clean week-over-week delta report you can paste into any client deck.
Quick start
| Tool | What it tracks | Link |
|---|---|---|
| AI Brand Visibility Tracker | Brand mentions, position score, share of voice | apify.com/dbott23/ai-brand-visibility-tracker |
| AI Citation Auditor | Domain citations, citation rank, cited URLs | apify.com/dbott23/ai-citation-auditor |
Both are pay-per-check — no subscription. First runs are free on Apify's free tier.
Questions? Drop them in the comments. I'm also tracking which queries and engines tend to surface brands most reliably — happy to share what I've found.
Top comments (0)