Apple Private Ad Measurement (PAM) – an early primer

February 21, 2024

What is PAM?

PAM builds upon previous Apple privacy measurement proposals such as “PCM” (Private Click Measurement) and “IPA” (Interoperable Private Attribution), as well as the Google Privacy Sandbox “ARA” (Attribution Reporting API) proposal.

PAM aims to “define a system that can measure the effectiveness of advertisements on the Web without tracking users across websites”.

Similar previous proposals from big tech such as Apple and Google have not yet reached broad consensus or adoption across major browsers. PAM differentiates in some ways with previous proposals by relying heavily on the “Prio” framework – a  “privacy-preserving system for the collection of aggregate statistics” developed by researchers at Stanford university in 2017 (ref: “Prio: Private, Robust, and Scalable Computation of Aggregate Statistics”).

Apple note that “Prio type aggregations standards are being drafted by the IETF Privacy Preserving Measurement Working Group. Using this type of underlying technology means that ad attribution will benefit from the ecosystem supporting a wide variety of measurements”.

Key features of the PAM proposal include:

  1. Storing data about ads displayed to a user on their device (e.g. browser). This data is fully manageable by the user, but is not readable or detectable by websites.
  2. Conversion information is similarly stored on the user information. This is then used to derive a “histogram” (similar to ARA) to in order to attribute a conversion with the matching ad impressions in that users device storage.
  3. “Differential Privacy” techniques are used to “add noise” to the histograms to protect individual user privacy.
  4. These histograms are then used to provide reports back to advertisers and publishers to attribute impressions and conversions.

How does PAM compare to similar proposals?

  • PCM & PAM: Rely on the user’s device to join an advertisement impression and conversion together. Data is stored on the browser instead of a central “party”.
  • IPA & PAM: Rely on Multi-Party Computation (MPC) frameworks to segment data across multiple partners so that no individual organisation can track an individual.
  • ARA & PAM: Provide measurement that ties together the “linkages” between advertisers and publishers” that are otherwise not trackable without 3rd party cookies.

PAM data & reporting

AdFixus Measure

AdFixus supports the development of new privacy-preserving alternatives to measurement and attribution. We understand that there is uncertainty in what the future of measurement in the open internet will look like and whether there will be standards that are accepted by the W3C to be adopted across all browsers.

We believe that the PAM proposal represents an important test for the future of standardisation across the internet for privacy-preserving measurement. How will major browsers collaborate and reach a consensus on a standard methodology?

Our mission at AdFixus is to evolve digital identity to bring balance between privacy and relevancy whilst giving individuals control over their data. Third-party measurement solutions have been increasingly limited in the last few years in their ability to provide useful insights in a privacy-first internet. We have been working with forward-thinking Publishers and Advertisers for several years to solve this issue through our patented first-party technology.

Whilst proposals like PAM help move the industry forward, high-performant organisations are not seeking to adopt any single proposal as a “silver bullet” solution to 3rd party cookie deprecation and usage for advertising, measurement, and user identification. AdFixus customers are adopting similar solutions by first defining a comprehensive strategy that supports their entire customer ecosystem. The AdFixus solution is a critical piece of infrastructure for our customers to execute a 1st party strategy that is:

  1. Supported by all browsers .
  2. Implemented without limitation on all owned domains.
  3. Owned fully without external 3rd party ownership.
  4. Comprehensive beyond cross-site cookie use-cases.
  5. Aligned to evolving privacy-legislation and browser privacy protocols.
  6. Durable across enterprises into the wider data sharing ecosystem.

Additional Information

PATCG?

The PATCG (Private Advertising Technology Community Group) is a W3C group chartered “to incubate web features and APIs that support advertising while acting in the interests of users, in particular providing strong privacy assurances”.

PATCG is an open online community group that anyone is able to freely join. Proposals that are submitted to PATCG are then evaluated and may be adopted as a work item to define a specification. Ultimately, the group seeks to identify potential standards that can be proposed to various standards bodies to be adopted across the internet.

Popular recent proposals that have been worked through the PATCG include the Topics API, IPA, and PCM.

Further Reading

  • Original proposal submitted to PATCG (https://github.com/patcg/proposals/issues/17)
  • Private Ad Measurement individual draft proposal (https://github.com/patcg-individual-drafts/private-ad-measurement).
  • Overview of PAM by Criteo Engineering (https://medium.com/criteo-engineering/apples-private-ad-measurement-52dcd41021f7).
  • Prio research paper by Stanford (https://crypto.stanford.edu/prio/).
By clicking “Accept All Cookies”, you agree to the storing of cookies on your device to enhance site navigation, analyze site usage, and assist in our marketing efforts. View our Privacy Policy for more information.