SKAN: A practical standard for the implementation of SKAdNetwork
Since we’ve announced our support for SKAdNetwork and released the open-source code for everyone to start exploring, we’ve received massive interest from the industry and had many conversations with top advertisers, publishers, ad networks and MMPs on its potential.
The industry, and we, are torn between accepting the likely loss of IDFA (that served so many positive and valuable use-cases) and getting ready for the new reality, one characterized with fragmented data, partial consent, fingerprinting, on-device solutions and … SKAdNetwork.
SKAdNetwork has some benefits, but it also has some severe limitations, and we are actively participating in the development of alternative privacy-friendly methods with other players, and we would love to see them succeed.
Still, with so many unanswered questions and a nearing deadline, there is a void created that is currently being filled with uncertainty and confusion.
We believe that SKAdNetwork – whether we like it or not – will be part of that new reality. And in its current form, SKAdNetwork is a bare-bones framework that requires careful implementation and responsible governance between multiple entities in the ecosystem to ensure that it’s practical and useful.
That’s why we created SKAN. To try and give us a path forward in a land of uncertainty.
Introducing SKAN: A practical model for SKAdNetwork
Today, we’re excited to announce SKAN – a proposal for a standard implementation of SKAdNetwork. SKAN describes how SKAdNetwork should be used in a standardized way by the ecosystem to overcome some of its limitations and enable its adoption in a scalable and trustworthy way.
This spec is the result of many conversations with folks across the industry about the challenges and issues SKAdNetwork will present.
With SKAN, we are aiming to achieve the following:
- Create a reference implementation plan for SKAdNetwork, to avoid fragmentation and a broken mobile advertising ecosystem.
- Design a way for SKAdNetwork to be compatible with parallel methods like IDFA-based attribution (when consent is given) and fingerprinting (when permitted) – both of which will be required.
- Provide a way to centralize and manage the scattered SKAdNetwork postbacks.
- Govern SKAdNetwork’s “conversion value” and define a standard implementation.
- Provide a solution for the fraud risk presented in SKAdNetwork (such as the conversion value manipulation, repeating transaction ID attacks, and geo manipulation).
- Define how marketers and ad networks can measure ROAS and other KPIs.
- Define the infrastructure necessary for centralized reporting for advertisers.
We invite everyone to join us and contribute to this spec. We’ve deliberately excluded our name, or any other company’s name from the spec, to encourage collaboration, suggestions, improvements, and most of all criticism.
We know that one player can not shape that future on their own. It’s like that old saying “it takes a village.” We will need to come together as an industry — advertisers, ad networks, publishers, and MMPs — to shape the future of marketing measurement… and we need to do it quickly with iOS 14 launching in September.
If you’d like to learn more, connect with other thought leaders, ask questions and participate in very lively discussions around IDFA and attribution privacy in general, we invite you to join the Mobile Attribution Privacy (MAP) Slack group: click here.