The honest one-paragraph version
Kantar Marketplace is Kantar's self-serve platform, operational across 70+ markets, delivering panel-driven research outputs in hours rather than days. AI is layered across the marketing lifecycle. It's the most credible incumbent platform in the "traditional firm going AI-native" lane.
Theia is structured intelligence built AI-native from the first commit. We integrate Demand, Visibility, Sales and Perception across Google, Amazon, AI Overviews, deep web and editorial — refreshed continuously, source-traceable, reproducible at run-id level. We don't operate a panel.
The two platforms answer overlapping but genuinely different questions. The right choice depends on whether your decision needs panel-validated stated preference or continuous structured-graph intelligence.
Where Kantar Marketplace is strong
- Panel access at scale — millions of pre-recruited respondents across 70+ markets
- Established methodology library — concept testing, brand tracking, ad testing, all with decades of validation
- Procurement-friendly — large enterprise sales relationship, easy approval through existing Kantar contracts
- AI bolt-on — GenAI tools across the marketing lifecycle, increasingly capable
If your decision needs stated-preference data from a representative panel, with the validation track record of Kantar's methodology, Kantar Marketplace is purpose-built.
Where Theia is built differently
| Dimension | Kantar Marketplace | Theia |
|---|---|---|
| DNA | Panel + AI overlay | AI-native, integrated graph |
| Primary signal | Stated preference (surveys, panels) | Revealed preference (search, review, transcript, citation) |
| Pillars covered | Brand / consumer perception primarily | Demand · Visibility · Sales · Perception (all four) |
| Cadence | Wave-based + on-demand projects | Continuous weekly refresh |
| Multi-language | Yes (translate surveys) | Native source-language extraction + harmonisation |
| AI Overview / LLM citation tracking | Limited | Weekly across ChatGPT, Perplexity, AIO, Claude |
| Deep web / B2B sources | Not in scope | 8,000+ classified deep-web sources (B2B pipeline) |
| MCP / natural language query | Through Kantar tooling | Direct MCP exposure |
| Reproducibility | Methodology-validated waves | Run-id reproducible at claim level |
| Strategy generation | Deliverable-led | L1-L4 strategy chain ship-ready |
Where the genuine overlap is
Both platforms produce brand and category intelligence. Both increasingly use LLMs. Both serve CMOs and insight directors.
The substantive difference:
- Kantar Marketplace tells you what a representative panel says when asked. Validated, statistically defensible, comparable across waves.
- Theia tells you what the open-web behaviour reveals. Continuous, multi-source, integrated with sales and visibility.
Stated preference and revealed preference are different evidence types. Both have a place. The gap between them is often the interesting story.
Should you have both?
For large CPG and consumer brands, yes — they're complements, not substitutes. Kantar Marketplace for the stated-preference wave; Theia for the continuous structured graph between waves. The combination covers what neither covers alone.
For mid-market consumer brands, the choice is real:
- Kantar Marketplace if procurement requires established-firm credibility and your decisions are driven by stated preference
- Theia if you need continuous monitoring across all four pillars and your decisions need source-traceable evidence
For B2B / industrial brands, Theia's deep-web pipeline is structurally beyond what Kantar Marketplace addresses. Industrial buyers don't sit on Kantar panels.
For research firms, the choice is starker: white-label Kantar Marketplace, or white-label Theia. Different commercial models, different scope.
Pricing comparison
- Kantar Marketplace — typical engagement £50-150k per wave, with subscription options for larger clients
- Theia Tier 3 (consumer brands) — £6k/month for 4 countries (£72k/year) including all four pillars + L1-L4 strategy
- Theia white-label (research firms) — £6k-12k/month with volume tiers unlocking flat-fee economics
Different models, different value propositions.
What we'd want a fair comparison to include
- Time-from-question to evidenced-answer for an open-ended category question
- Multilingual capability (native vs translated)
- AI Overview citation tracking
- Reproducibility of a specific finding across a 6-month gap
- Source traceability per claim
- Continuous refresh vs wave cadence
We're happy to run that test side-by-side for prospective clients. The fair comparison isn't credentials; it's whether the platform answers the actual decision question on time, with evidence, in a way the team can defend.