Next Advertising Era - Google Privacy Sandbox: Attribution Reporting API
What is Chrome Ad Measurement, how does it work, and what does the future of measurement hold for us?
Not everyone's caught on that third-party cookies are not just the foundation of the advertising targeting infrastructure, but they're also the driving force behind measurement infrastructure: without third-party cookies, even today, it's not possible to measure post-view conversions, while for post-click ones, the Ad Tech industry is now leveraging first-party cookies.
The industry needs to come up with a new measurement infrastructure. Right now, there are two solutions on the market:
The industry needs to come up with a new measurement infrastructure. Right now, there are two solutions on the market:
- Apple's SKAdNetwork and Private Click Measurement, both active on the iOS platform, iPadOS, and Safari browser across all of Apple's operating systems.
- Google is testing the Attribution Reporting API for both Android environments and the web, with different variations, but with the same basic principles in common.
What's the Attribution Reporting API?
The Attribution Reporting API is a measurement framework that doesn't allow linking click, impression, and conversion events to a single user.
The infrastructure is built on two approaches:
- Event-level reports: These attribute single clicks or impressions to a conversion, but the conversion-level info is super limited. The fewer conversions you have, the less detail you get. For instance, we might know which group a conversion was attributed to, but not which ad creative or keyword. As the number of conversions goes up, so does the detail level, but there's a catch: noise is added to the data by sending random info for 5% of the events to protect user privacy.
- Summary reports: These give you really precise conversion-level data, but you can't link it to a specific click or impression event. The framework uses techniques to keep the user's privacy intact.
Why have two measurement methodologies been introduced?
Measuring the performance of an advertising campaign has different goals, and they're not all as obvious as we might think. From a marketer's perspective, there's a need to optimize each placement in a campaign and to know which and how many conversions they've generated. But the same set of information can be used by the advertising platform to spot fraud attempts. That's why event-level reporting was introduced, to show data like this
The click 200400600 made by the user Bob_doe on the site news.example has resulted in a transaction on the site shop.example.
Event-level reporting can't help understand the return on investment of a campaign because it lacks cost detail, and primarily because, by design, the infrastructure introduces noise: random data. Summary reports, thanks to precise data, allow the analysis of campaign's performance.
The campaign 1234567 on news.example led to 518 conversions on shoes.example with a spend of $38,174. Half of those conversions came from NYC, USA
Conclusions
Measurement will be possible with noise, maybe more than we see today. Right now, we're oblivious to the noise and trust the tools we have, under the illusion that everything is accurate. In the future, we'll be aware that these measurements are scientifically imperfect and we'll need to work hard to improve the output. This might not be such a bad deal for us industry insiders.
Support from other browsers hasn't been announced yet. I believe Microsoft and the Mozilla Foundation are watching how things evolve, with the Redmond company potentially moving quicker than the Red Fox. Apple will develop its own technology, but we're starting to see common ground emerging between the two frameworks. Let's see how the next version of SKAdNetwork will evolve.
Previous Episode of: Next Advertising Era
Reading Suggestions
Be Data Literate by Jordan Morrow that is driving my professional choice
Present Beyond Measure: Design, Visualize, and Deliver Data Stories That Inspire Action by Lea Pica that is supporting my data storytelling
IAB Tech Lab Identity Solution guide from IAB to understand Alternative ID
Marketing in the messy middle Google's document to guide us through the Messy Middle, Mountain View's perspective on summarising the user's decision-making process.