Cookies remain on Chrome, but Safari, Firefox and iOS already block them — which leaves roughly 40%+ of the traffic an SSP monetizes with no third-party identifier attached. Buyers price what they cannot address at a discount, and that discount lands on your publishers. Cookieless Audience gives SSPs a way to close the gap: pre-computed audience attributes for 102 million domains, aligned with IAB Audience Taxonomy 1.1, plus a real-time API for page-level audience segmentation. Use it to power seller-defined audiences, curation products and inventory packages that describe who the audience is — without a single identifier or any PII.
An SSP sees domains and URLs on every request, but domains and URLs carry no audience meaning by themselves. On the cookieless share of traffic — all of Safari, Firefox and iOS — there is no third-party identifier to fall back on, so audience-driven budgets skip that supply or bid it like blind reach.
Domain-level audience data reverses the direction of the problem. Instead of asking “who is this user?”, it answers “who reads this property?” — with coded demographics, 285 sub-interests, 283 purchase-intent segments, B2B firmographics and 1,667 deterministic personas per domain, each attribute carrying a banded confidence score. Because the vocabularies are aligned with IAB Audience Taxonomy 1.1, the same attributes crosswalk directly into the segtax declarations and deal packages buyers already understand.
For large multi-topic publishers in your supply, the real-time API extends the same vocabularies to individual URLs, so a section — not just a site — can carry its own audience description.
// One supply domain, as the database describes it { "domain": "example-fitness-site.com", "age_bracket": "25_34", "gender_skew": "balanced", "income_level": "upper_middle", "interests": [ "INT.healthy_living.fitness_and_exercise", "INT.healthy_living.nutrition" ], "purchase_intent": [ "PI.recreation_fitness.gyms_and_health_clubs", "PI.sporting_goods.exercise_and_fitness_equipment" ], "confidence": "high" } // Every value comes from fixed v1.0 vocabularies // aligned with IAB Audience Taxonomy 1.1.
The same dataset feeds three sell-side products. Each one turns raw supply into something a buyer can plan against and pay a premium for.
Most publishers in any SSP’s supply have no first-party data to declare. With domain-level attributes for every property, the platform can construct taxonomy-aligned SDA segments on their behalf and emit them under segtax: 4 — the full workflow is on the seller-defined audiences page.
Filter the whole supply footprint by persona, interest, intent, demographics or firmographics, and package the qualifying domains behind a Deal ID: “fitness enthusiasts, 25–44”, “B2B software evaluators”, “new parents, upper-middle income”. The mechanics are covered under inventory curation.
Attach audience attributes to marketplace listings, packages and publisher one-sheets so supply teams sell audiences, not URL lists. Confidence bands tell your team — and the buyer — how firmly each claim is supported by evidence.
The dataset joins on the one key an SSP always has — the domain. Everything downstream runs in your own systems.
Match the licensed domain file against your active supply list — the domains you see in ads.txt, sellers.json and your own inventory tables.
Keep attributes banded medium or high. Packages built only on well-supported attributes survive buyer scrutiny.
Express each product as boolean logic over coded fields — interests, intent, demographics, firmographics, personas.
Generate the qualifying domain set per segment, attach it to Deal IDs, and crosswalk attributes to Audience Taxonomy 1.1 nodes for SDA declarations.
Quarterly refreshes re-score the corpus; fixed v1.0 vocabularies mean your segment logic keeps working unchanged. Use the per-URL API for section-level splits in between.
A mid-size SSP wants a health-and-fitness audience package it can offer to sports-nutrition and activewear buyers across its whole supply footprint — including the long-tail sites that have never had first-party data. The segment logic below runs against the licensed domain file.
| Segment rule (coded values) | Reads as | Role in the package |
|---|---|---|
INT.healthy_living.fitness_and_exercise | Fitness and exercise interest | Core inclusion rule — defines the audience theme |
PI.sporting_goods.exercise_and_fitness_equipment or PI.recreation_fitness.gyms_and_health_clubs | In-market for fitness equipment or gym memberships | Commercial layer — what makes the package worth a premium |
age_bracket ∈ {25_34, 35_44} | Readership skews 25–44 | Matches the buyer brief’s demographic target |
confidence = high on the interest attribute | Strongly supported classification | Quality gate — keeps weak matches out of the deal |
| Alternative IDs / identity graphs | Contextual signals | Domain-level audience data (this dataset) | |
|---|---|---|---|
| Works on cookieless traffic | Only where the ID is present (login, consent) | Yes, everywhere | Yes, everywhere — keyed on the domain |
| What it describes | An individual user | The page’s content | The property’s audience — demographics, interests, intent, personas |
| Long-tail coverage | Thin outside logged-in media | Good, but audience-blind | 102M domains, head to tail |
| Privacy posture | Consent-dependent identifiers | No PII | No PII, no identifiers, fixed public vocabularies |
| Sell-side product fit | Buyer-driven; SSP is a conduit | Adjacency packages | SDA, curation deals, audience-described packages — and it composes with both others |
No. There are no identifiers and no PII anywhere in the pipeline. The dataset describes properties — who reads a domain, expressed in fixed, versioned vocabularies aligned with IAB Audience Taxonomy 1.1 — rather than resolving individual users. That is exactly why it keeps working on the 40%+ of traffic where Safari, Firefox and iOS block third-party cookies.
No. The audience attributes are pre-computed for 102 million domains, so even a long-tail publisher with no login, no CDP and no analytics integration arrives with a complete coded profile. Publishers that do have first-party data can layer it on top; the domain-level layer gives every property in your supply a consistent baseline.
You filter domains on coded attributes, crosswalk the retained attributes to their IAB Audience Taxonomy 1.1 nodes, and configure your exchange or wrapper to emit the node IDs in OpenRTB user.data with ext.segtax: 4. Because the vocabularies were built aligned with that taxonomy, the crosswalk is a lookup, not a modeling project. The step-by-step version is on the seller-defined audiences use-case page.
No, and we deliberately do not claim impression-level pre-bid classification. The API provides per-URL audience segmentation for planning and analysis — profiling sections of large publishers, auditing sample URLs from a package, building section-level curation. Bid-time decisions remain in your own systems, informed by the packages and declarations you built from the data.
The segtax-4 declaration workflow in full detail.
Turning audience attributes into Deal IDs.
The same substrate, from the curator’s seat.
How the buy side plans against the same data.
Look up any domain in the audience demo, or license the domain file and ship your first taxonomy-aligned supply packages this quarter.
Open the audience demo See database pricing