Segmentation

Audiences built on where people actually go.

Ready-to-activate segments grounded in real-world behavior, not demographic guesswork. Off the shelf or built to your brief, and delivered into the platforms you already run.

Get in touch with us

City map with three red circular zones marked with burger, car, and dumbbell icons and white streets.

Real behavior.
Ready to activate.

Audiences is the segmentation layer: it turns the foundational mobility signal into segments you can run on any channel. Three things make a Dymos audience different.

Built on real behavior

Every audience is grounded in observed real-world visits, where people genuinely went, not demographic guesswork or fading cookies.

Any channel, any identifier

The channel sets the format: MAID for in-app, residential IP for CTV and desktop, HEM when you start from CRM. One audience layer, delivered the way each platform needs it.

Sized to activate

We link and enrich identifiers, not identified people. A residential IP identifies a household context, not a person, and location resolves to an inferred area, never a precise address. You get an addressable, segmentable view of what kind of place a signal comes from, without dossiers on who lives there.

Segmentation

How it works

From real-world presence to a segment you can run:

Visits come first: Raw presence is modeled into real visits to brands, categories and points of interest.
Segments map to a taxonomy: Visits are classified by brand, category or custom list, kept separate from the audience name, so segments can be created or changed without recomputing visits.
The channel sets the identifier: MAID for in-app, residential IP for CTV and desktop, HEM when you start from CRM.
Built for scale: Segments are sized to spend and optimize, and refreshed regularly.

Map with icons for restaurant, fitness center, and car showing clusters of customer data points.

Three audience families

One audience layer, three families. Pick the family that matches your goal, and the identifier that matches your channel.

Map showing fast food, fitness center audiences with visitor clusters and nearby grocery store and car dealership.

Visit audiences

Reach people by where they actually went. Our visit modeling separates genuine visitors from passers-by, so a segment reflects people who actually stopped, and can be tuned toward those likely to have bought.

Granular Audience

Visitors of specific brands, categories and points of interest, across 8,600+ brands and 150+ categories.

Real-World Behavior

The behavioral core of the offering: reach people by where they actually went.

Activation

Use it for conquest, retargeting, retail visitation and store-visit planning.

Reach the people who actually showed up, not the ones a model guessed at.

Map showing two audience zones: Paris pet lovers with 25,000 households, and H3 cell fine dining area.

Location and proximity audiences

Audiences defined by where people are based: home or main area, postal code, H3 cell, or a custom geography you draw yourself, from store catchments to business districts and event venues.

Target by where people genuinely are, down to the area that matters.

Profile audiences

Affinities, purchasing-power proxies and territorial profiles, built from real-world behavior rather than declared data. Example segments range from Gym Goers and Pizza Lovers to Frequent Air Travelers and Museum Visitors, available in major markets. Useful for planning, audience qualification and lookalike modeling, across MAID, IP and HEM.

Add the why behind the where, so a segment explains interest, not just location.

Put to work

What teams do with it

Across programmatic activation, conquest, out-of-home and measurement, the need is the same: audiences grounded in where people actually go, broad enough to perform, and available without building your own pipeline first.

When you need to activate quickly across more identifiers and broader geographies, without standing up your own location pipeline first.

When demographic or modeled data is not enough, and you want audiences grounded in where people actually went.

When you run conquest, category-expansion or trade-area strategies that generic data cannot support.

When you plan, measure or extend out-of-home and DOOH campaigns, and need to know which screens reach which audiences, and when.

When you need audiences sized to spend and optimize, not just to sample.

Use cases

Audiences in action

Real challenges, solved with audiences grounded in real-world behavior.

Use case

IP audiences for CTV and audio

A media player used residential IP as its key identifier across CTV, audio and web, channels where device IDs are scarce or absent, so IP was how it recognized and reached audiences. But it could only target with its own first-party data, which capped both the scale and the precision of what it could offer. It needed to extend targeting beyond that first-party base without taking on a device-ID dependency those channels could not support.

Audience performance was comparable to the client's own first-party data, a strong result for third-party enrichment, and enough to move the relationship toward a long-term partnership.

Use case

Conquest campaign for a retail challenger

A retail brand wanted to win shoppers from a dominant competitor, but its targeting reached broad demographics with no way to focus on people who had actually visited rival stores.

Spend concentrated on genuinely contestable shoppers, lifting store-visit rates and turning a broad awareness push into a measurable footfall play.

Use case

Planning and measuring DOOH with real-world audiences

A digital out-of-home player had a list of available screens but no way to know which screens reached which audiences, at what times, or to prove and extend a campaign's impact.

Screen and time-slot selection lined up with real audience presence, out-of-home exposure could be tied to in-store visits, and exposed audiences were extended into digital channels