“Data-Driven Thinking" is written by members of the media community and contains fresh ideas on the digital revolution in media.
Today’s column is written by Hugo Loriot, managing director at 55.
Are the recently announced Chrome privacy controls a good answer to the latest versions of Apple’s Intelligent Tracking Prevention (ITP)?
Much has been said about them, whether it is to welcome a careful move with little expected impact on the industry or blame Google for prioritizing advertisers over end users.
Regardless of whether you think Apple is far too restrictive or Google way too lax, digital marketers and their partners now must deal with a two-speed ecosystem – one that creates inconsistent datasets and apples-to-bananas comparisons that lead to the wrong marketing decisions.
To recap, with ITP 2.2, Apple now automatically blocks third-party cookies, which are used for ad frequency management, retargeting and view-through attribution modeling. It also limits the lifespan of most first-party tracking cookies to either one or seven days.
With the new Chrome privacy controls, Google will let developers define whether the cookies they set can be used in a third-party context. Users will be offered an easy way to block third-party cookies on an opt-in basis.
How will these policies play out? Consider a brand marketer who wants to measure the impact of a campaign on ecommerce revenue, leveraging a best-of-breed, fully integrated tech stack. Say a user is exposed to a display ad on day one of the campaign, visits the website directly for the first time on day two, and then converts after clicking a branded paid search ad on day eight.
On Safari, where view-through attribution is made impossible and clickstream data perishes after one or seven days, the journey would be broken and channel performance would be reported as follows:
- Display: one impression, no assisted visit or assisted conversion
- Direct: one new visit, no last click or assisted conversion
- Paid search: one new visit, one last click conversion
On Chrome, where only opt-in users block third-party cookies and web analytics cookies are not impacted, the journey would look much different:
- Display: one impression, one assisted conversion
- Direct: one new visit, no last-click conversion, one assisted conversion
- Paid search: one returning visit, one last-click conversion
The information available to the marketer is incorrect on Safari and correct on Chrome, but the main issue is that the marketer will probably not be able to determine which dataset to make marketing decisions on, because the data is skewed in up to half the population.
Unfortunately, the validity of the data in this case does not matter, because nothing can be done about it. ITP is here to stay, view-through attribution on Safari is dead, and it’s probably not a good idea to re-engineer all your web analytics cookies to circumvent the one- or seven-day expiration policy. Google may or may not enforce a stricter policy, and it’s not in anyone’s interest to play guesswork.
However, what is in everyone’s interest is to know how the data is collected, the technical constraints of each environment, and the risks of aggregating metrics across Safari and Chrome.
A marketer must know when and how to segment reports per browser or operating system, whether or not it is relevant to extrapolate users’ behavior on Chrome to other browsers, and why an audience segment’s size differs between a first-party-cookie-led web analytics solution and a DSP.
Campaign Manager and Google Ads have started providing baked-in conversion estimates to help marketers better assess their campaign performance in a cross-device and cookie-less world. But media plans are still built with the assumption that retargeting can help reach all past visitors, ad frequency will be incremented by one every time the same user is exposed to an ad, and exposures can lead to conversions. This is (often) true on Chrome, but it is certainly not on Safari, and the sooner marketers take that into account, the better.
The question is not whether Apple is killing the industry or if Google has a good reason for saving it, but how to avoid making unfortunate marketing decisions by looking at meaningless aggregates.