The advertising industry has reached a critical juncture with cookie deprecation testing in Chrome. Google’s Privacy Sandbox aims to reduce cross-site and cross-app tracking while keeping online content and services free for all. Sounds idealistic – because it is.
Google has invited players from across the ecosystem to test the Sandbox and its privacy-preserving APIs to assess if the platform is a viable alternative for advertising at scale. Early adopters are now well past 30 business days of Privacy Sandbox testing. And, as with all routine testing, there are expected bugs to work out.
The IAB Tech Lab’s Privacy Sandbox Task Force has shared its case study analysis, and the report findings were skeptical at best, offering more questions than answers.
And Mediavine’s own testing reveals that the Privacy Sandbox’s Protected Audiences API (PAAPI) causes increased latency, decreasing viewability and yield.
Given these issues, publishers simply cannot afford to test PAAPI at scale.
Latency’s drag on viewability
At present, when PAAPI wins an auction, the result is an average increase of 1500 milliseconds of additional latency.
Publishers aim to reduce latency as slow delivery results in lost impressions, lost revenue, nonviewable ads and a poor consumer experience. Our testing results average a viewability rate of 39% according to GAM when PAAPI wins the auction. That’s far below the industry minimum standard of 70%. Our educated hypothesis is that latency is a direct contributor to low viewability, which advertisers will not value.
In its current state, Google is recommending a traditional server-side auction followed by a Privacy Sandbox auction, nearly doubling the average auction time. During this time frame, publishers like Mediavine set variable timeouts between 1600 and 2300 milliseconds. Adding a sequential auction increases latency dramatically, resulting in a direct threat to revenue.
Best practices have been recommended by Google’s Privacy Sandbox team, including the request for buyers to limit interest group owners and bidding, reusing bidding scripts, monitoring auctions and using timeouts and limits. The only issue is that the auction does not account for any publisher controls or preferences.
Publishers need the freedom to run auctions asynchronously because the current version of three auctions in succession breaks functional ad-loading experiences, including lazy loading.
This high latency will majorly impact the probability that an ad will be displayed at all and massively devalue available inventory. Google needs to share controls with publishers or the testing can’t move forward at scale.
Publishers have important choices to make
We know the current publisher experience in the Privacy Sandbox is not contemplated, but that’s a solvable dilemma.
We, as an industry, have talked about this problem ad nauseam. The early results are leading to a long list of complaints and questions. But those of us who are actively participating in testing can do something about it.
If testing is a threat to monetization, so is the product as a whole. We need to make that clear.
Mediavine has shared all of our results with the Privacy Sandbox team. We recognize PAAPI is not an end solution but a work in progress that requires feedback. Publishers need more controls and reporting capabilities within the Sandbox now.
Currently, publishers face restricted alternatives. One option is to make a huge investment in becoming a top level seller, forgoing any AdX demand. The other is to continue to benefit from AdX but resign all semblance of control and reporting in PAAPI. This ultimatum creates an undue hardship for publishers. We hope that providing feedback at this stage gets us to a place where we can continue to test and work with our partners on solutions.
Where does the Sandbox go from here?
Google has been clear. The Privacy Sandbox is “not designed to offer 1:1 replacements for third-party cookies or cross-site identifiers.”
The industry needs to work together to make significant improvements to user privacy. A viable alternative for advertising dollars is better than no alternative.
Here are some recommendations for the industry to take immediate action:
- The time is now, even if it’s uncomfortable. If you’re not currently testing the Sandbox or one of the related APIs, you’re lagging behind your partners and competitors.
- Share your results. The more often the buy and sell sides have an opportunity to align on questions and asks for Google, the better our chances of improving the current solutions. The Privacy Sandbox team, IAB Tech Lab and the CMA are all invested in this analysis.
- Continue to test despite setbacks. Remember, the present discoveries represent preliminary findings and do not constitute a definitive analysis of the offering. This is one step in the journey to privacy-centric advertising.
- Be prepared to wait. Even with a proactive approach to testing and feedback cycles, there will be a delay. Privacy-centric programmatic advertising was not built in a day.
The deprecation of third-party cookies is undoubtedly a transformative moment for digital advertising. Privacy Sandbox may not be the solution we all expected it to be, but there’s only one way to find that out – rigorous testing.
In doing so, the industry can not only adapt but take some control over the cookieless future, turning this challenge into an opportunity for improved advertising experiences.
“The Sell Sider” is a column written by the sell side of the digital media community.
Follow Mediavine and AdExchanger on LinkedIn.
For more articles featuring Amanda Martin, click here.
