01
How many
The size of any segment: a market, a category, a brand, or people who did A and B but not C — in a second, with the threshold shown.
People who visited three or more competing brands this month in Mexico
Behavioural market intelligence · web + app · aggregates only
Measure the market, not the model
Omnimetra measures digital markets from the real behaviour of millions of people: how many of them are in your market, who they are, where to find them and how it changes. One panel, web and apps together, an answer in a second, a passport on every number.
Data passport
- period
- 2026-09
- panel builds
- 283
- taxonomy
- v2.3
- method
- panel IDs, web + app
- disclosure
- ≥ 100 IDs per cell
- reproducible
- yes, by build id
- ~1 bnpanel rows a month
- web + appin one visitor ID
- 36 moof history on one taxonomy
- < 1 sto size a narrow segment
- ≥ 100 IDsdisclosure threshold
Four questions every market team pays for
Not page views. Not a search index. Not a survey. The measured behaviour of a panel, answered with a passport.
02
Who
The profile of that segment against any base: interests, channels, device, attribute buckets, with lift, never a person.
How the audience of a new entrant differs from the market
03
Where
The sites, apps and platforms where the segment concentrates — the ones that bring new reach and the ones that bring the same people again.
Placements by lift for lapsed customers of a category
04
What changes
Shares, loyalty, switching, cohorts and seasonality over 36 months on one taxonomy, with the data version on every chart.
Who gained after a regulatory change and from whom
Real behaviour, not an estimate
Panel IDs from data clean rooms
Web visits and app usage with a persistent pseudonymous visitor ID, processed inside clean rooms. We never receive personal data; the platform only ever shows aggregates above a disclosure threshold.
Web and apps in one ID
The same visitor carries a web and a mobile mark, so overlaps, loyalty and switching between a site and an app are measured, not inferred.
A passport on every number
Period, data build, taxonomy version, method and threshold travel with every figure. Re-run the query on the same build and the number is the same — in a deck, in a report, in a regulator's file.
How it works
- 01
Clean-room panels
Partners deliver monthly batches of pseudonymous visits for web and apps. Attribute buckets arrive by our specification; people never leave the clean room.
- 02
A labelled entity graph
Every site and app is labelled by country and vertical from registries, our own crawl, AI labelling and organisation enrichment, then linked site → brand → operator.
- 03
Answers with passports
Markets, segments, overlaps, cohorts and media plans are computed on precomputed sets, with the disclosure policy applied before anything leaves the engine.
One platform, four ways to buy
Start with a market, grow to segments and media planning, add custom research or licence the data itself.
Platform
One vertical, up to three countries: market shares, brands, trends, overlaps, monthly report with passport, alerts.
from $2,000 / month →
Platform Extended
Segment Studio, cohorts and switching, media planning, attribute buckets, all countries of the vertical, 36 months, Ask AI.
from $6,000 / month →
Custom research
One question → a report with a passport in 2–4 weeks: market entry, segment size, channelization, competitor audience.
from $7,500 / study →
Data access
Data API, aggregate feeds by brand and country, BI exports, Snowflake and AWS listings, MCP. 30-day data trial.
from $10,000 / month →
Deep in your industry, not generic
Each vertical is a pack: its taxonomy, brands, licences, events and segment templates. Open one, get the market.
iGaming & betting
Shares of operators by country, cross-sell sport → casino, mirrors and licences, effect of regulation.
Fintech & banking
Audiences of banks, neobanks, payment and investment apps; overlap, switching and web-to-app migration.
Crypto
Exchange and wallet audiences by country, switching after incidents, overlap with fintech and trading.
E-commerce & retail
Share of shoppers across marketplaces and retailers, loyalty and exclusivity, sales-day cohorts.
Media & streaming
Unique reach and duplication across streaming, publishers and social; release cohorts and switching.
Travel & mobility
OTAs, airlines, hotels, rail and ride-hailing: seasonality, cross-brand shopping, direct vs meta.
Gaming & esports
Mobile titles, stores and esports media: cross-play between titles, publisher portfolios, launch cohorts.
Telecom & super-apps
Operator apps, ISPs and super-apps: audiences by country, bundle partners, switching.
Why not a traffic model, a search index or a survey?
Every method answers something. Only measured panel behaviour answers overlap, loyalty and switching across web and apps.
| Omnimetra | Modelled traffic | Search index | Surveys | |
|---|---|---|---|---|
| Basis | Measured panel behaviour, web + app | Statistical model of traffic | Search demand | Self-reported answers |
| Overlap of any number of brands | Exact, from IDs | Up to a few sites, estimated | No | Declared, small samples |
| Loyalty, switching, cohorts | Yes, 36 months | Limited | No | Rarely |
| Narrow segment in a second | Yes, with passport | No | No | No |
| Apps and web together | One ID | Separate estimates | No | Declared |
| Personal data exposure | None, aggregates only | None | None | Consent-based |
Omnimetra
See your market in a 30-minute walkthrough
Bring one question. We answer it live on real data for your country and category, with the passport on screen.
Questions
What is behavioural market intelligence?
Measuring a digital market from the observed behaviour of a panel of real people — which sites and apps they use, how often, and in which combinations — instead of modelling traffic or asking people what they do. Omnimetra reports only aggregates above a disclosure threshold.
Where does the data come from?
From partner data clean rooms that process web visits and app usage with a persistent pseudonymous visitor ID. Attribute buckets are returned by our specification. No personal data reaches the platform.
How is this different from Similarweb or a search index?
Modelled traffic estimates visits per site; a search index measures demand for brand names. Omnimetra measures what the same people actually do across sites and apps, so overlaps, loyalty, switching and narrow segments are exact for the panel and carry a passport.
Can we activate audiences?
Only through a partner clean room, only where the country × vertical × action is allowed by our compliance matrix, and only above the minimum size. The platform itself never exports IDs.