That is according to a survey (Apple, IDFA and iOS14: New Challenges, New Opportunities for Marketers) recently conducted by AppsFlyer and MMA Global. The low confidence in SKAdNetwork is primarily due to its lack of measurement granularity. Of course, this isn't a surprise to us, as we have described in our earlier posts. The other key findings were:
- 74% of respondents agree that publishers will suffer revenue losses
- 63% of marketers will invest in probabilistic modeling to prepare for post-IDFA
- Audience targeting/remarketing will take the biggest hit (as shown in the chart below)
Now, what should YOU do?
Firstly, measurement granularity is a major gap within SKAdNetwork and we're glad to hear that marketers are now acutely aware of this limitation. If you are one of those marketers that are somewhat-to-very-likely to adopt SKAdNetwork, we believe that SKAdNetwork is a good starting point but make sure that you clearly understand its glaring gaps:
- SKAdNetwork will leave marketers with only 100 campaign IDs of granularity
- It does NOT offer a way to cohort KPIs by install date
- Critical post-install performance KPIs that are foundational to your mobile growth strategy today like retention, revenue, LTV, and ROAS, will disappear and be replaced by a single conversion value with its own significant limitations.
Secondly, as you can see from the survey there is a lot at stake (74% agree that publishers will suffer revenue losses). Time is running out and the competition is breathing down your neck. Why not make sure your post-IDFA measurement plan is foolproof? You can do that by adding our top-down incrementality measurement solution which can co-exist with SKAdNetwork, and offers a comprehensive, granular and accurate alternative that addresses the above glaring issues with SKAdNetwork. Remember, there is a lot at stake. Contact us today to discuss and close any gaps in your IDFA preparedness.
Thirdly, the focus on probabilistic modeling appears to be short-sighted as Apple has already stated in their current terms that 'fingerprinting' will not be allowed. More explicitly, in the FAQ section of that document, Apple has the following question and answer:
Can I fingerprint or use signals from the device to try to identify the device or a user?
No. Per the Developer Program License Agreement, you may not derive data from a device for the purpose of uniquely identifying it.
Instead of focusing on 'fingerprinting' or probabilistic modeling, apply your precious cycles to quickly testing our top-down incremental approach which is the ONLY future-proof measurement solution available today. It requires really low effort on your part and we invite you to test drive it for FREE.
Please submit your questions below. We look forward to answering them and making the future of your measurement brighter!
