November 15, 2019
Incrementality Testing: What You Should Know About This New Marketing Tactic
Update: See our most recent ITP blog post for the latest developments.
Advertisers are beginning to see the impacts of Apple’s ITP 2.1 on tracking and measurement across analytics, marketing automation, and CRM platforms, with the update shrinking remarketing audiences and lowering conversion reporting for ad campaigns. On the eve of Apple’s ITP 2.2 rolling out soon, which likely will have even more significant effects on performance tracking in Google Analytics, it is important to project the degree of the impact to your data and adjust goals accordingly.
Without the cookies that tell Google Analytics where a converting browser came to your site from, Google is going to attribute those conversions to a different channel source. So, if you stick to analyzing performance for each channel the way you always have, that will lead to undervaluing the performance of your marketing in various channels. What can you do about it? The simplest thing to do is to adjust paid media goals that are measured in Google Analytics or other analytics platforms. Not doing so will likely result in decreased growth and missed business opportunities. It’s not the actual paid media performance that is going down, it’s just the optics of that performance in Google Analytics.
We’ve outlined how to assess the impact and think about new goals below, so that you can prepare your stakeholders for how marketing reporting, analysis, and decision-making will need to change.
The degree of impact for ITP 2.1 will be tied to your ratio of Safari traffic, as well as the proportion of conversions that occur after 7 days.
The first step is to determine what percentage of your current traffic and revenue is coming from Safari users.
The second step is to determine what percentage of current Safari traffic has upgraded to ITP 2.1. There are several different ways to do this in Google Analytics, but we recommend creating a new Segment to monitor this traffic.
Note: This particular Segment will slightly overcount current Safari ITP 2.1 adopters by 1 to 2% since it captures a small number of Safari (in-app) browser users who have yet to update to the latest browser version. That is of little concern since these Safari users will eventually upgrade to ITP 2.1 and beyond.
The third step is to determine how much of your revenue (or how many conversions) come in after 7 days. ITP 2.1 cuts off tracking cookies after 7 days.
Now it’s time to do some math. Here is an example for an ecommerce website:
What does this mean? If you set channel goals using Google Analytics data, you should reset your goals with the above percentages in mind. In fact, with ITP 2.2 coming soon, you should calculate the impact for all Safari conversions/revenue happening after 1 day:
In this example, once ITP 2.2 is fully rolled out, you should consider lowering your channel goals by 26% to account for the Safari conversions/revenue after 1 day that are no longer being attributed to the appropriate channel in Google Analytics.
Will this be a significant change for advertisers who have relied on Google Analytics data to set goals? Yes. Is it the end of the world? No. Google Analytics has never been a perfect platform, and there have always been serious flaws with Google Analytics data. From cross-device tracking to lack of impression reporting to heavily undervaluing Facebook ads, Google Analytics is imperfect and shouldn’t be used as a singular source of truth. Instead, Google Analytics data should be treated as a piece of the puzzle.
Speaking of the puzzle, it’s about to get more complicated. There are rumors of the Chrome browser considering similar steps to ITP, and those are looking more real with comments on cookies from a Google executive at a recent conference. [Update: Google formally announced changes coming to the Chrome browser at their Google I/O event]. In addition, there has been little clarity provided from digital ad platforms about the current and future impacts of ITP (not to mention Chrome). Tracked conversions and retargeting audience sizes are already seeing impacts. Everyone from Facebook to Pinterest to LinkedIn to marketing automation platforms should be affected, but they haven’t been saying much.
In the short term, platforms like Google Ads might use modeling to account for non-tracked Safari conversions, but there are larger changes afoot. All advertisers will have to adapt in the near future. Whether that means planning for a post-cookie world or doing more with incrementality measurement, advertisers may need to get used to having less tracking info at their fingertips. The good news is that the value of the right message, at the right time, to the right audience isn’t changing (even if how the data is tracked and reported on does change). For more on this topic as it develops, sign up for our email newsletter using the widget to the right of this post.