Query fan-out:
See what AI searches before it answers.
Vercite captures fan-out queries from ChatGPT and Perplexity for your tracked prompts. It also collects People also ask questions shown beside Google AI Overviews. See which searches name your brand, which name competitors, and how the wording changes over time.
7-day free trial · cancel anytime
Available on Starter, Professional
Of the queries and questions naming a tracked brand: Nike, both, or only a competitor (adidas, ASICS, Brooks).
ChatGPT48 searches52%46%
Perplexity64 searches11%42%47%
AI Overview33 searches18%9.1%73%
76 queries and questions look for a competitor without Nike.
Brands the engines put in their searches, by engine.
- On109
- ASICS86
- Nike69
- Brooks67
- HOKA66
- Saucony46
- New Balance43
- adidas35
- Salomon35
- BOA23
Nike tracked competitorTop 10 of the brands named.
Real Vercite data · Nike demo account · last 30 days · highlight a share, switch engines and brands
Trusted by leading brands





What can you see behind each tracked prompt?
You can inspect captured fan-out queries and People also ask questions, with brand names matched and results split by engine.
Fan-out queries tied to prompts
See the background searches captured with each ChatGPT answer, and with each Perplexity answer when Perplexity reports them. Keep the connection to the original prompt, so a search string has the buying question that prompted it beside it.
People also ask beside AI Overviews
Read the People also ask questions captured alongside Google AI Overviews. Use them as a separate source of question wording when planning what a page should answer.
Share of brand searches
For each engine, compare the share of captured searches naming your brand only, your brand alongside a competitor, or only a competitor. This measures brand names in searches, separately from mentions in the final answer.
Brands and organisations named
See which names appear in the captured searches and questions, split by engine. Check whether a topic repeatedly includes a competitor before you decide which examples or comparisons to investigate.
A searchable explorer
Search the collected wording across your tracked prompts. Look for a product type, brand name or phrase, then inspect the prompts and engines associated with the matching searches.
A choice of periods
Review the last week, month or 3 months. Compare the language appearing in your captured searches over time, with the available history kept in mind for each engine.
How does Vercite capture query fan-out?
Vercite stores the fan-out queries captured with tracked answers and matches brand names in their wording.
- 01
Track your customer prompts
Add the prompts you want to measure through prompt tracking. Use the questions customers ask about products, comparisons and buying decisions, so the searches stay tied to relevant topics.
- 02
Capture the searches and questions
Each tracked run stores ChatGPT’s background searches. For Perplexity, it stores the searches Perplexity reports, or otherwise the follow-up questions it suggests. Google AI Overview runs also collect the People also ask questions shown alongside the result.
- 03
Match the brand names
Vercite cleans the captured text and matches the brands and organisations named. The engine breakdown lets you inspect where your name appears alone, alongside competitors or nowhere in the captured wording.
- 04
Inspect the gaps and recurring wording
Open the overview or search the explorer for a topic. Read the searches naming only competitors, note recurring questions, and bring the original wording into your content briefs.
Fan-out queries and People also ask across the covered platforms
Vercite captures fan-out queries from ChatGPT and Perplexity, plus People also ask questions alongside Google AI Overviews. The Google questions keep their People also ask label throughout the product.
- ChatGPT
- Perplexity
- Google AI Overviews
Query fan-out use cases
Find searches that name only competitors
Inspect competitor-only wording to identify comparisons and topics where your brand is absent from the captured searches.
Plan content around observed searches
Use the phrases engines search for to investigate whether your pages cover the product details and comparisons appearing there.
Spot changes in language
Compare the prompt with its captured fan-out queries to see when local-language questions lead to English searches.
Give writers real questions
Bring People also ask wording into a brief, then use it to assess which reader questions the page answers.
Does ChatGPT search for what you asked?
ChatGPT’s fan-out queries often use substantially different wording from the original prompt.
Across 17,806 captured ChatGPT fan-out queries, 64% of the words never appeared in the prompt, and only 0.5% of searches were phrased as questions. 41% used a different language from the prompt, mostly English. The study found about 1.9 fan-out queries per answer and about 38 distinct searches across roughly 25 repeat checks of the same prompt.
Questions about Query fan-out
Related features
Start seeing the searches behind your tracked AI answers.
7-day free trial · cancel anytime.