How does AI search work for your category?
AI tools break every buyer question into smaller questions before they go looking. That set of smaller questions is different for every industry: different constraints, different buyer signals, different tools dominating each stage. This page maps how getting cited works in nine categories and shows the modeled breakdown for each.
Your buyer asks one question. The AI asks eight. You are missing seven.
This isn't SEO. The AI picks passages, not pages. The selection happens out of sight, before your buyer sees any answer.
The head query arrives
A buyer types one question. The AI treats it as a retrieval task, not a search. The two are different: a search engine ranks pages; an AI breaks the question apart and builds an answer from pieces.
Fan-out happens invisibly
Before any retrieval happens, the AI expands the one question into 6 to 18 smaller questions. The buyer never sees these. They carry the intent, the constraints, who's asking, and the gaps the buyer didn't know to ask about. Each smaller question gets its own search.
Passages are picked, not ranked
For each smaller question the AI holds several candidate passages at once and picks one. This is the selection event. Winning it isn't about keyword density. It's about carrying the right evidence, scope, and structure for that specific question.
The answer is stitched from winners
The AI stitches the chosen passages into one answer. Citations point at the source of the winning passage. A brand quoted in three of eight passages shows up as a top recommendation. A brand missing from all eight is invisible, no matter how big or well-known.
Every AI rewrites differently
ChatGPT strips constraints and generalizes. Perplexity keeps the wording close. Copilot compresses to a few words. Claude applies an honesty filter. Google AI Mode expands with structure. The same question produces a different set of smaller questions on each. One content plan doesn't cover them all.
Citabld maps the full space
We work out the complete set of smaller questions for your category, across all nine AI tools and your buyer types, then write the content that wins the selection events that matter most.
The smaller questions aren't universal.
A "best X software" question in fintech generates compliance-gated smaller questions. The same question in logistics generates location-anchored and integration-specific ones. The same question in EdTech forks into completely different paths depending on whether the buyer is an IT director or an individual teacher. Your industry shapes the whole breakdown.
Price limits often get changed by the AI; compliance limits almost never do. Your industry decides which constraints survive.
Some industries have one buyer type. Others have four. Each one produces a different set of questions from the same starting question.
High-stakes industries trigger proof-seeking questions automatically. The AI asks for evidence before the buyer does.
Perplexity dominates fintech research. Google AI Mode dominates local and retail. The same brand needs different coverage on different tools.
B2B SaaS comparison questions get generated heavily. Service businesses generate fewer comparisons and more proof-seeking ones.
Real estate and logistics questions almost always keep a location through rewrites. Most other categories drop it.
Select your category.
Modeled rows for each industry. Every row is MODELED, derived from our method, not captured from live tool logs. The paid library adds measured rows where available.
B2B SaaS
The category every AI tool knows best and defaults to least honestly.
Buyers look at 4 to 7 tools using AI before they talk to sales. The AI runs comparison, alternatives, and fit-checking questions all at once. A brand missing from the comparison layer loses the shortlist before the demo.
best crm for mid-market sales teams
salesforce alternatives under $100 per user
hubspot vs pipedrive for 50-person team
crm with built-in revenue forecasting 2026
what crm do series b startups use
crm that integrates with slack and notion
is salesforce worth it for small business
crm migration downtime risk
These rows are MODELED. They come from our decomposition method, not from live tool logs. Google's surfaces share no query data with anyone. The paid library adds MEASURED rows from the tools that expose them.
See the modeled fan-out for your head query in 90 seconds.
Free, no account required. We work out the smaller questions for your query and write one sample content piece for the one you choose.