Step01
Paste a prompt your customers ask AI.
The same prompts you'd want to rank for in Google still apply. "Best CRM for SaaS." "How to switch from Mailchimp." Anything commercial.
A research tool by New Chemistry
Query Fan-Out ToolSee what LLMs actually search for your industry.
Every prompt an AI receives gets fanned out into a dozen sub-queries it runs against the live web. Those sub-queries decide which sites get cited. Paste a prompt. We'll show you the fan-out, the cited sources, and a priority score on every one.
Try
how does GEO differ from SEO
best B2B SaaS marketing agencies
Notion alternatives for engineering teams
Free, no credit card required
What it does
01
Map the fan-out
One prompt becomes 8 to 12 sub-queries. We surface them, classified by category and intent so you can see what the LLM is really looking for.
Score every sub-query
Volume, keyword difficulty, AI Overview presence, and a priority verdict on each. Skip the dead ends. Build for the wins.
See the cited sources
Live web results pulled for every sub-query, with the domains AI Overviews already cite. Know exactly who you're up against and where to insert yourself.
The concept
02
What actually happens when you ask an AI a question.
An LLM doesn't have a magical index of the web. It takes your prompt, breaks it into a handful of related sub-queries, and fires those at search engines like Google in parallel.
Then it pulls the top results from each one, finds passages that agree, and synthesizes a final answer. The whole dance happens in seconds, and the user never sees the searches.
GEO strategies are shifting constantly, but optimizing for query fan-out topics has been an effective strategy and helps you prioritize traditional search and AI search at the same time.
How it works
03
From paste to playbook in 30 seconds.
Step02
We fan it out the way an LLM would.
Built on Google's published methodology. We generate the categorized sub-queries, then run live web searches on each to surface real cited pages.
Step03
You get a verdict on every sub-query.
Volume, keyword difficulty, whether an AI Overview already shows up, and a worth-building / watch / skip score so you know what to do.
What you get
04
Categorized fan-out
See every sub-query, classified with consistent methodology.
Related, comparative, entity expansion, reformulation, recent, implicit. Six categories pulled from the published methodology, so you can spot which angles you're missing without guessing.
- 6 primary categories
- Intent type per query
- Plain-English rationale
Sub-query · 02Worth building
Sub-query 02
Worth building
Salesforce alternatives for venture-backed SaaS companies
comparative
commercial
Sub-query 03
Watch
HubSpot vs Pipedrive for sales teams under 20 reps
comparative
informational
Priority scoringper query
Volume2,400
Difficulty42
Commercial valueHigh
AI OverviewYes
Mid-volume comparative query with an AI Overview already firing. Five domains in the cited set, none of them G2 or Forbes. Solid build target.
Verdict scoring
A score on every sub-query, so you know where to invest.
We pull volume, KD, CPC, and AI Overview presence from real search data, then label each query worth building, watch, or skip. No more guessing which terms are crowded versus winnable.
- Live Semrush data
- AI Overview check
- Worth-build verdict
Cited sources
The actual pages an LLM would pull from, for every sub-query.
Real web results, ranked by what the engine surfaces today. Use them to spot the listicles you should be in, the comparison pages that need to mention you, and the gaps where a fresh page can rank fast.
- Live web results
- AI Overview domains
- Publication dates
Top cited03 sources
-
Best CRM Software for SaaS Companies in 2026 g2.com/categories/crm-saas
-
10 Salesforce Alternatives for Modern Sales Teams attio.com/blog/salesforce-alternatives
-
SaaS CRM Buyer's Guide: What to Look For saastr.com/saas-crm-buyers-guide
AI's canonical answergenerated
For most SaaS startups, HubSpot, Attio, and Pipedrive are the strongest contenders[1].
HubSpot wins for teams that need marketing, sales, and service in one suite. Attio is the modern pick for venture-backed teams who want a data model that fits product-led motions[2]. Pipedrive remains the cheapest credible option under $20 per seat[3].
Salesforce only pays off above ~50 reps or in regulated industries.
Canonical answer
See what the AI would actually answer, in full.
The synthesized response an answer engine would write from the cited pages. Numbered citations linked to the source list. This is the answer your prospects are reading.
- Full synthesized answer
- Inline citations
- Mirrors AI Overview style
An example, end to end
05
Here's what came back for "best B2B SaaS marketing agencies".
A real fan-out from a real prompt. Two of the seven sub-queries scored worth building. One was already cited by AI Overviews.
best B2B SaaS marketing agencies
Sub-queries7
Unique sources31
Worth building02
Sub-query 01
Worth building
top B2B SaaS marketing agencies for product-led growth
commercial
Sub-query 02
Watch
SaaS demand-gen agency vs in-house team
comparative
informational
Sub-query 03
Worth building
best content marketing agencies for B2B SaaS 2026
commercial
Sub-query 04
Skip
how much do SaaS marketing agencies cost
reformulation
informational
Run your own
Free, no credit card required
The playbook
06
You have your fan-out. Here's the right play for each sub-query.
Your move depends on two things: are you ranking, and how competitive is the term? Four scenarios, four plays.
01
You're in top 10
Don't rebuild. Tighten.
You're in the mix. Add the exact sub-query phrasing as a section heading. Make sure the answer lives in a clean 150 to 300 word passage right under it. LLMs grab passages, not pages.
Quick win · 1 hour
02
Pages 2-5
Foundation, not focus.
Either expand the existing page with a section that answers the sub-query head-on, or split off a new page if it's meaningfully different. Internal link from the parent using the sub-query as anchor text.
Mid-effort · 1 sprint
03
Not ranking, brutal competition
Skip the head-on fight.
Get mentioned in the listicles, review sites, and round-ups that already rank. Submit your site, reach out to the author, get added to existing top-ranked content. LLMs cite those listicles. Now your name is in the answer.
Highest leverage · ongoing
04
Not ranking, low competition
Build the page. This is the goldmine.
Use the sub-query as your H1. Answer it directly in the first paragraph. Include a comparison table, a clean definition, and at least one concrete number. LLMs love uncompetitive sub-queries because there's not much else to choose from.
Most underrated · build now
Universal habits
07
A few habits that lift every page, no matter the scenario.
Use the sub-query phrasing exactly.
LLMs match on semantic similarity, and exact phrasing is the highest-confidence match. Lift the words from the fan-out into your H2 and first sentence.
Write in passages, not walls of text.
Each section answers one sub-query in 150 to 300 words. That's the chunk size LLMs prefer to lift directly into an answer.
Add data and specifics.
"Sales rose" is forgettable. "Sales rose 34% to $4.2M in Q3" is the kind of sentence LLMs pull verbatim. Numbers, dates, named brands, all win.
Spread your message off your own site, too.
LinkedIn, guest articles, podcast transcripts, Reddit threads. The more places your message lives, the more raffle tickets you have to be the cited source.
Questions
08
Your questions, answered.
If you have a question we missed, the team at New Chemistry will happily field it.
How is this different from regular keyword research?
Keyword research tells you what humans type into Google. Query fan out tells you what an LLM types into Google after it reads what a human asked it. Different layer, different leverage. You should do both.
Do all LLMs fan out the same way?
No. Each model has its own reasoning chain and its own training data, so the same prompt fans out differently in ChatGPT, Claude, Perplexity, and Gemini. The terms you see here are directional. The patterns hold across models even when the exact phrasing differs.
How often should I re-run a fan-out?
Any time your prompt's context shifts. New competitor in the market, a product launch, or a new year. The tool injects today's date into the fan out, so a "best X" prompt run in 2026 will look different than the same prompt run in 2025.
Why are some of my sub-queries about a totally different industry than I expected?
That's a real product insight, not a bug. Lots of acronyms and short brand names have dominant meanings outside your industry. If your fan-out comes back full of results from a completely different field, the LLM is telling you another meaning of that term wins on the open web.
Trying to rank for the bare term without context is a bad bet. Use a longer phrase that includes your industry, or invest in a clear disambiguation page that defines the term in your context.
What happens to my email if I unlock the full report?
You get one email a week from New Chemistry on AI search visibility. No drip, no sales touches, no list trades. Unsubscribe in one click. We don't share or sell your address.
What is this built on?
The methodology comes from Google's published patent US20240289407A1 ("Search with Stateful Chat"). Live web results come from a real-time SERP layer. Volume, KD, and AI Overview signals come from Semrush. We pay for the API calls so you don't have to.
Built by New Chemistry
Prefer to hand it off?
If turning the fan-out into pages that rank isn't how you want to spend your week, that's what we do.