Google started deprecating third-party cookies on January 4, 2024, but most companies are unprepared for the cookieless future.
A 2023 EMARKETER survey reported that 50 – 90% of industries hadn’t yet invested in non-cookie programmatic ads. Now, feedback on Google’s Privacy Sandbox fuels more confusion. The IAB Tech Lab found few use cases viable in the Sandbox. The UK’s Competition and Markets Authority (CMA) raised similar concerns, spurring about a 60-day standstill period.
So, if you’re an advertiser, what should you do next?
With no one-size-fits-all solution, marketers need a strategic approach. Despite all the doom-and-gloom news out there, cookieless marketing will be better, and there are practical strategies advertisers can implement now to prepare for third-party cookie deprecation.
Cookie Deprecation = A Good Thing?
Soon, many solutions advertisers take for granted won’t be feasible. But does this mean the world will end?
No. We’ll move towards better marketing overall by moving away from third-party cookies as the main foundation of digital media targeting and measurement.
For years, third-party cookies were like a placebo: They felt like they were working, but what’s better than placebos? The real thing. Third-party cookies are a brittle and unreliable data source loosely connected across platforms to infer behavior and identity. The generated profiles are a bit like horoscopes — some will be accurate, but many segments will be wrong. Human behavior isn’t consistent.
Instead, by relying on first-party data, marketers can advertise to the actual people they intend to. This data will be more accurate and precise for targeting and measurement. We’ll be able to align metrics with business outcomes beyond on-site behaviors and measure true business lift.
No-Bull Testing Strategies
Here are some strategies to put into practice starting now.
1. Set Up a Cross-Functional Task Force
Assemble an organizational team to test cookieless identity, targeting and measurement solutions. Involve people from media buying, analytics, partnerships and a tech expert who can evaluate realistically what works and what doesn’t. You’ll also want buy-in from leadership, especially for changes you’ll need to make post-testing.
2. Expect Some Smoke and Mirrors
A lot of what different platforms are saying they can do in a post-cookie landscape is most likely impossible — and when it is possible, it’s dubious how they can get there. Many present different options as infallible solutions, but I believe it’s all going to explode very soon. There isn’t a replacement for third-party cookies. So, expect some smoke and mirrors when meeting with vendors. Don’t take things at face value; push back against what they say if you have doubts, and do your homework to see through the bull.
3. Create Intentional Testing Frameworks
Design testing frameworks so that each test is specific and actionable. Your testing framework should answer the following questions:
- What are we trying to learn?
- How will we approach this test?
- What does success look like?
- What is the specific hypothesis?
- How will we activate on the learnings?
The timeframe for each test will vary depending on how long it’ll take to reach statistical significance.
4. Test Multiple Solutions
In a cookieless ecosystem, everything that’s out there will be like slices of Swiss cheese. While one solution may deliver some outcomes, there will still be holes. Advertisers will need to essentially layer the pieces of Swiss cheese on top of each other so there are fewer holes. In short, we’ll need different solutions to fill the gaps third-party cookies have filled. Therefore, you’ll need to test various vendors simultaneously.
For identity solutions, you’ll need to test deterministic and probabilistic solutions providers. Deterministic providers are identity graphs that make unique, consistent IDs based on email addresses given with consent. Top ones to test include The Trade Desk’s Unified ID Solution 2.0 (UID2), LiveRamp ATS and IntentIQ. Probabilistic providers create modeled IDs algorithmically by analyzing various signals including a user’s IP address, browser, screen solution and more. Test solutions like ID5, LiveRamp RampID and Lotame Panorama ID.
When it comes to targeting, you’ll want to look at targeting against a deterministic, authenticated ID solution like UID2, then augment with something probabilistic. Or, you can use blended solutions like RampID or ID5. You’ll also want to revisit contextual targeting since contextual targeting doesn’t rely on cookies. Grapeshot, Peer39 and GumGum are good providers to test.
Measurement solutions to test should fall into three camps: deterministic, probabilistic and decentralized. For deterministic solutions (i.e., providers that create pure streams of authenticated-only, logged-in traffic from publishers’ deterministic IDs), test the measurement capabilities of UID2, LiveRamp and IntentIQ. For probabilistic solutions, test media mix modeling solutions, whether your own or open-source solutions like Meta’s Robyn or Google’s LightweightMMM, and incrementality (both forensically and synthetically tested). Finally, some decentralized measurement options to test include the Google Privacy Sandbox Attribution Reporting API, SKAdNetwork and the Interoperable Private Attribution (IPA).
The beauty of the solutions listed above is that they focus on everything real — people and business outcomes — instead of low-value conversion pixel data.
5. Know the Roles Deterministic and Probabilistic Solutions Will Play
In a perfect world, all IDs would be deterministic. But we don’t live in a perfect world. The reality is, for most advertisers, deterministic IDs will only make up a very small portion of the user base they’re targeting — most likely only 20 – 40% of your buy. Your best bet is to build a deterministic foundation and then set up a probabilistic series of solutions for the other 60 – 80%.
6. Follow Privacy Sandbox Guardrails
Despite the concerns the ad tech space has brought up about the Privacy Sandbox, you should do your own testing to find if they align with what your tests show and see how your measurement will be affected after third-party cookies go away.
As mentioned, definitely test the Attribution Reporting API. Also, test the Google Topics API. Go into the tests being aware of the limitations. For example, both APIs will deliver delayed data, whether 24 hours or 7 days after, depending on the API. Know that it’ll be harder to manage your budget and optimize your campaigns on a day-to-day basis.
Go Fearlessly Into a Cookieless Future
By following these tips and strategies, you’ll be well on your way to thriving in marketing’s new era. Like any good journey, there will be hiccups and hangups along the way, but in the end, you’ll discover how to navigate the post-cookie landscape with the right solutions in place. You’ll also have confidence that your data is more accurate and will measure the real impact of your campaigns on bottom-line results.