Publisher provider signals (PPS) allow buyers to receive audience and contextual data from publishers. Publishers are able to communicate their audience attributes, such as behavioral and interest-based data, and contextual data to programmatic bidders by using standardized taxonomies:
- IAB Audience Taxonomy version 1.1 (includes Demographic, Interest and Purchase Intent categories) provides the industry with a common nomenclature for audience segment names and improves comparability of data across different providers. Download IAB Audience Taxonomy.
- IAB Content Taxonomy version 2.2 ( including video signals like CTV context) provides the industry with a common nomenclature for contextual segment names. It categorizes page content without revealing user information. Download IAB Content Taxonomy.
Some benefits of PPS:
- PPS filters, cleans and standardizes publisher 1P data to industry-aligned taxonomies to make it simple and easy for bidders to ingest and map to their own taxonomies
- PPS addresses the need for personalized ads and privacy so that publishers don’t need to share user identifiers.
- Publishers can create valuable audiences that buyers and advertisers can target without having to expose user-level data.
- PPS makes audience buying more accessible to buyers today in the absence of third-party cookies in browsers like Safari, and in environments like apps and OTT/CTV where there are limited contextual signals available in the bidstream.
- PPS enables buyers to reach data cohorts across different publishers, multiple browsers and devices.
How does PPS work?
At a high level, PPS works as follows:- A user visits a publisher’s property, including website, mobile app or connected TV app.
- The publisher generates signals that they expect buyers will find valuable to reach their advertising client’s goals. Examples could include:
- Audience segment (for example, Interest, In-Market), inferred based on user activity on their properties (web traversal history, forms, purchase history, and so forth).
- Content signals, particularly valuable when content is not crawlable (app, video, web content behind login).
- Using tools within Google Ad Manager, the publisher standardizes the content / audience signals into one of the PPS supported taxonomies (for example, IAB content 2.2).
- The publisher sends an ad request to Ad Manager, with content and audience signals sent as taxonomy category IDs inside a predefined key-value.
- Ad Manager sends these signals to programmatic bidders.
- Advertisers can target just as they do today.
Buyer Integration
New fields related to the PPS signal provider, taxonomies and PPS segments will be available in the RTB callout:
We use data objects to send these signals to allow bidders to better differentiate when values are declared and sent by a publisher as first-party data.
PPS uses the following OpenRTBBidRequestfields
for sending audience signals:
BidRequest.user.data.name
BidRequest.user.data.segment
BidRequest.user.data.segment.id
BidRequest.user.data.ext
BidRequest.user.data.ext.segtax
PPS uses the following OpenRTBBidRequest
fields for sending content signals such as genre:
BidRequest.content.data.name
BidRequest.content.data.segment
BidRequest.content.data.segment.id
BidRequest.content.data.ext
BidRequest.content.data.ext.segtax
Bidders who already support the IAB’s Seller Defined Audiences (SDA), and who receive Ad Manager bid requests via OpenRTB are able to ingest and use PPS because it uses the same fields as SDA (for both audience and contextual signals).
Bidders can reference the Seller Defined Audiences specs for IAB.
Video & Audio signals (Beta)
While content and audience taxonomies supported above are available for all types of inventory such as web, app, and video, there are also additional signals on the RTB callout that are supported by PPS. These are:
Content rating
- Represents the content rating of the underlying video and audio content stream
- Supported values are G, PG, Teen and Mature.
Content.contentrating
Audio feed type
- Represents different types of audio content.
- Supported values are Music, Radio and Podcast.
Audio.feed
Production Quality
- Represents the content quality of the video & audio content stream.
- Supported values are Professionally Produced, Prosumer, and User Generated.
Video/audio.prodq
Content delivery
- Represents the delivery of video or audio content stream.
- Supported values are Streaming, Progressive, and Download.
Video/audio.delivery