If You’re Not Testing SKAdNetwork Right Now, Your Activity Is at Risk

By Ran Milo

There are situations in which “wait and see” is a legitimate – prudent, even – strategy. Craving a first-generation tech product? Maybe you should wait and let others test out the kinks. Planning a vacation abroad this summer? Yeah, things are looking up, but maybe don’t become too invested just yet.

Then there are the situations in which waiting and seeing is really not a good option. And you probably already know where we’re going with this: the importance of having a working SKAdNetwork implementation right now cannot be overstated for anyone whose fortunes are even somewhat dependent on iOS marketing.

The problem is massive, gargantuan even. If you don’t think it is, we urge you to dig deeper and understand just how much the entire iOS marketing industry revolves around IDFAs right now. And not just the workflows, but the tools as well. How many BI systems do you know that don’t build their information stack from the user level up? And what do you think will happen to this foundation when the user-level data is gone, and everyone’s IDFA is either zeroes or just doesn’t exist?

If Your Insights Vault Is Empty, You’re in Trouble

The disconnect between upper and lower funnel marketing data is among the most severe implications of Apple’s ATT framework. Without dual IDFA opt-in, the connective tissue between what’s happening outside your product and within it is gone. Yes, there’s SKAdNetwork, but its single conversion value and anonymization is hardly comparable to the data-rich environment that fuels marketing campaigns these days.

Obviously, when there’s a disconnect between the two parts of the funnel, using lower-funnel data to optimize marketing becomes impossible. And this is where your insights vault comes into play. Right now, before iOS 14.5 is out, you have the very unique opportunity to receive two streams of data: IDFA-based deterministic data, as well as what you’ll receive from iOS 14.5 onwards.

Now is your only opportunity to see deterministic data side-by-side with SKAdNetwork data

By looking at the two streams at the same time, you can try and find correlations. Which iOS 14.5 metrics have meaningful correlations with your KPIs, and which do not? What are your most important conversion events? These analyses will fill your insights vault, which you will open on the day IDFA-based marketing is no longer an option.

At Bidalgo, we use the ads’ creative assets, and their aggregated measurements, as a privacy-forward connector between upper funnel and lower funnel data. This approach helps our clients optimize their activity without any user-level data collection.


Trust, But Verify

Before you begin assembling your insights, though, you need to understand if your SKAdNetwork implementation actually provides you with the correct data. And this, again, is much easier when you have something to compare it to.

Here’s what happened with a client of ours: their SKAdNetwork implementation showed similar install numbers, with discrepancies easily-explained by the time frame-based anonymization Apple uses, but conversion data was very different. Were they to wait for iOS 14.5 before testing this out, they couldn’t have detected the issues as quickly.

Big discrepancies? There’s an issue somewhere.


They’re not the only ones who had implementation hiccups, around 40% of our clients’ initial SDKAdNetwork campaigns had worrying data discrepancies. The one thing that’s common to all these hiccups is the value of having deterministic data as a validator for SKAdNetwork data. Let’s remember that everything conversion-related in SKAdNetwork is really complex. If you map the funnel events incorrectly, there will be no easy way to see this without having access to deterministic data. And yes, of course, our over-time graphs help here as well.

100% of the clients we spoke to say that seeing their SKAdNetwork implementation data side by side with deterministic data helped them prepare for what’s to come, including implementing product changes. For example, we see that SKAdNetwork reporting doesn’t support conversions after 7-day trials well, so if the marketing of your product relies on this specific event you might want to reconsider your onboarding flow.

Please, Act Now

Hopefully, we have convinced you to act now, whether on your own or by letting us help. And when you’re done with your SKAdNetwork implementation, we’ll be more than happy to talk to you about how you should refocus your optimization around creative assets. Our arguments are pretty convincing, we think: from not having to rely on user-level data at all, to future-proofing your activity against automation in bidding and audience matching. You can read more about our creative-centric approach here.


Published on March 29, 2021
Written by
Ran Milo

VP Marketing @ Bidalgo.

Get Started With Bidalgo Today