The premise
When you sell market intelligence platforms, the most credible thing you can do is run your own engine on your own category. So we did.
This post is the abridged public version of an internal discovery study we ran on "AI market research" — the category Theia sells into. Same methodology as every client engagement. Same source set (open web, editorial, comparative reviews, academic). Same outputs (demand pockets, brand-vs-market gap, content roadmap).
The findings shaped our website content over the last 48 hours. They might also tell you something useful about where the category is going.
The 7-cluster shape
The "AI market research" category resolves into seven distinct sub-categories that buyers, editors and search engines treat as related but separate:
| Cluster | What buyers ask | Dominant players |
|---|---|---|
| A | "Best AI market research platform" | Quantilope, Listenlabs, Attest, H-in-Q, Segwise — listicle-led |
| B | "LLM brand monitoring" / "AI Overview citation tracking" | Evertune, OtterlyAI, Authoritas, BrandLight, Nightwatch, Signal AI |
| C | "GEO / AEO / generative engine optimization" | Jasper, EnrichLabs, Search Engine Land discourse |
| D | "Synthetic respondents / AI personas" — and the critique | Simile ($100M raised), Quantilope, Listenlabs · critique: NIQ, Bellomy, VerianGroup |
| E | "Continuous market intelligence vs quarterly waves" | Luth Research (Q&A SEO), Conveo, GWI |
| F | "AI due diligence / PE portfolio monitoring" | Third Bridge, BDO, Brownloop |
| G | "B2B competitive intelligence / B2B AI visibility" | Alpha-Sense, Tami.ai, 2X (AI Visibility Index) |
Theia touches all seven. No single competitor we identified does. That's the strategic shape: the category is fragmented along these seven lines, and the integrated play is open.
The three biggest gaps
Running Theia's own brand-vs-market gap methodology on Theia's positioning revealed three UNTAPPED pockets — where the market is loud and our brand is quiet:
01 — Deterministic harness + audit trail
The EU AI Act high-risk deadline lands on 2 August 2026. Enterprise procurement is starting to ask "show me reproducibility, show me audit trail." We ship both in production — and we hadn't written a single page about it.
Glossary entries added: deterministic pipeline harness, reproducible AI research.
02 — GEO / AEO / LLM citation strategy
The fastest-growing category in the discourse, with Gartner forecasting 50% decline in conventional search traffic by 2028. Evertune dominates the citation-tracking conversation; nobody connects citation to broader strategy.
Pages added: /geo-aeo, /vs/brandlight.
03 — Synthetic respondents critique
The clearest critique discourse in the category — NIQ, Bellomy, VerianGroup have all published — but no published positive alternative. We are the alternative (revealed preference, native extraction, structured graph). We hadn't framed it that way.
Pages added: /vs/simile, /ai-market-research (with the "why we don't ship synthetic respondents" section).
The 11 unowned questions
From the discovery, eleven specific questions emerged as high-intent + low-quality-answer pairs — the kind of question that converts well in search and gets cited in LLM responses.
- What is AI market research? — Generic answers dominate. Theia now answers it at /ai-market-research.
- Are AI agents deterministic for market research? — Engineering blogs answer; nobody answers from the market research angle. We do at reproducible AI research.
- Can synthetic respondents replace real consumer research? — The critique is published; the positive alternative isn't. We claim it.
- How do I track LLM citation share for my brand? — Evertune owns this; we participate via GEO/AEO.
- GEO vs AEO vs SEO? — Acronym confusion; we clarify it.
- Why are 96% of B2B brands invisible in AI? — 2X published the report; the answer is deep-web coverage, which we ship for Canon B2B.
- What's the audit trail for AI-generated market research? — Almost nothing credible answers this. We do at deterministic harness.
- Continuous vs quarterly market research? — Luth dominates the Q&A; we extend with continuous monitoring.
- What is structured market intelligence? — We coined the category; we own it at structured market intelligence.
- How does AI-native diligence work in PE? — Third Bridge dominates; we answer with the Anchorage and Principality format.
- How do I white-label AI research for my agency? — Veridata and Stitch dominate the white-label search but they're survey-led. We pitch the AI-native white label at /for/research-firms.
What this method does for clients
The exercise we just ran on ourselves takes a Theia engagement about 5-7 business days for a client category. The output:
- 5-15 demand pockets, named and sized
- 5-10 brand-vs-market gaps, classified
- Competitive landscape with named players per lane
- 8-15 prioritised content actions
- Source-cited evidence on every claim
For a Tier 3 consumer brand engagement (£6k/month × 4 countries), this kind of discovery study is one of several monthly deliverables. For a one-shot diligence engagement (£45k, 4 weeks) it's the headline output.
For a research firm white-labelling Theia, this is exactly what you'd deliver to your end client — except your name is on the deck.
The two questions worth ending on
Q1: Why publish this? Because the methodology is the marketing. If we can produce a publishable category study on ourselves in an afternoon using the same engine we sell, that demonstrates the platform better than any feature slide. The risk is that competitors learn from it. The bigger risk is that prospects don't see what the engine can actually do.
Q2: What would your category study say? Run the same exercise on yours, and the answers will surprise you in three places: where you've been overselling messages the market doesn't care about, where you've been silent on properties the market loves, and which one or two questions in your category would convert if you just answered them properly.
This is what a 30-minute walkthrough at theia-digital.co starts with.
Coming next
Next post: Why we don't ship synthetic respondents — the structured critique of a $100M+ funded category, and the revealed-preference alternative.
Theia builds continuous structured market intelligence for consumer brands, industrial B2B, research firms, and PE. The full internal version of this study is in the content_plan folder of our public repo; this is the abridged Substack edition.