Skip to content
API Reference
Dataset Types
.md

Dataset Types

Read-only catalogs, one group per dataset type, that back the audience builder. Read these first to collect the ids and values the Audiences create body expects: IAB ids, CTV names, app bundle ids, cohort ids, POI ids, and so on. Catalogs shared across every dataset type - geography, languages, signal providers, endpoint connections - live in Common.

Every reference read is authenticated and JSON-only. Send Authorization: Bearer <token> and Accept: application/json on every call. All reference reads use the read rate bucket (120 requests/min per caller).

Two response envelopes are used:

  • Flat list reads return an array directly under data.
  • Paginated reads return data.items plus a data.pagination block (current_page, per_page, total, last_page). Paginated reads accept page, per_page, and search query parameters. per_page defaults to 500 and is capped at 500 on the Common, Web, and CTV reads. The four paginated Apps reads default to 250 and are capped at 500. A page above 10,000 is read as 10,000 (see Limits & Quotas).

Every list read - flat or paginated - accepts an optional search query parameter: a case-insensitive contains-match on the item’s display label.

  • Web Domain Visitors - IAB categories and subcategories, web domains, referrer domains, browsers, device types, makes and OSes.
  • Connected TV - CTV vendors, content types, content genres, channel names, device types, makes and OSes, connection types, ISPs and series.
  • Apps - app categories, tags, operating systems, bundle ids and taxonomies.
  • Cohorts - your company’s completed cohorts.
  • POI - POI segments, categories, brands and locations to pick from while building an audience (segments -> categories -> brands -> locations cascade), and the attributes scheduled at locations with their values (attributes -> attribute-values cascade). Managing the POI data itself is the separate My POI Data surface.
  • Transactions - affinity purchase categories, sub-categories and brands (categories -> subcategories -> brands cascade) plus the income, age, gender and ethnicity dictionaries: the values an AffinityTransactions dataset and a cross purchase analysis are scoped to.
  • Demographics - the gender, age range, marital status and income range dictionaries a Demographics dataset is scoped to.
  • Deidentified - the signal fields, in their six groups, a Deidentified dataset can carry into its delivery.
  • Profile Attributes - the attribute categories, keys and values a ProfileAttributes dataset is scoped to (categories -> keys -> values cascade), plus the delivered quarter window its date range is bounded to.
  • Origin has no catalog of its own: an Origin dataset is scoped with the Common geography reads - countries (with ?datasetType=Origin for the covered set), states, cities, DMAs and ZIP codes. Its fields are on Create Audience.

Company-scoped reads (cohorts, endpoint connections, your company’s partners) return your company’s records; credentials and partner secrets are never exposed.