Content
Stay up to date on the latest happenings in digital marketing
It’s 2025. What’s working for iOS mobile measurement? What is going to give you the data you need to beat the competition?
Sometimes it seems easier to identify what’s not working.
- SKAN is still hard
- IDFAs are still scarce
- Creative testing is harder
- Probabilistic has its own challenges
- And modeling … each major platform has released or is working on its own particular flavor of modeling iOS app advertising results, but it’s not always clear what goes into that
That’s why I recently discussed the future of iOS mobile measurement and marketing in 2025 with Jesse Lempiainen, co-founder of GeekLab, and Neils Beenen from Singular. (This is part of our 5-part Hack Gaming Growth event: see all the episodes in 1 place here.)
Hit play and keep scrolling for the highlights:
iOS mobile measurement: what’s working
There is good news. iOS mobile measurement still works … just not quite like it did, and not quite to the same level as it did. But there are developments that are hugely successful in restoring ad tracking, creative measurement, and conversion analysis.
Measurement methodologies that matter include:
- CAPIs (conversion APIs)
CAPIs give you reliable deterministic data on conversions: usually conversions on a website. Snap, Google, Pinterest, TikTok, and Meta all have CAPIs. Learn more about CAPIs here … - AEM (Meta) and Advanced Conversions (Snap) and Ads Conversion Modeling (Google)
Platform and ad network modeling give you additional data on ad impact, but they have some challenges (see below). - Advanced SAN from TikTok and similar programs from other platforms
Advanced SAN and similar programs provide more data for MMPs like Singular to make accurate attribution decisions, while remaining privacy safe. - SKAN/AAK
Yes, SKAN and its new iteration, AdAttributionKit are imprecise, partial, and delayed, but they are also deterministic and have value, even if they are not enough on their own. - Incrementality
There are many ways to test for incrementality (see a 2025 guide to incrementality here) but they all have 1 goal: is my ad spend on Network X or Platform Y incremental. In other words, does it move the needle on growth, or does nothing change if I stop my spend there? - Probabilistic
Let’s be honest, probabilistic is often used by programmatic networks, and likely others as well. - MMM (media mix modeling, or marketing mix modeling)
MMM can be hard to implement, though it has its benefits. Marketers are moving more to incrementality testing than MMM. - Data governance
Data governance is simply enforced naming conventions for creatives and campaigns — Singular offers a platform which automates the process — that unlocks otherwise-hidden creative analysis capabilities.
The challenge for 2025 is for marketers to combine multiple attribution methods to navigate iOS mobile measurement in 2025. (Without going insane, or trying to boil the ocean, or stressing too much when varying means of measurement return slightly different results.)
The good news is that Singular does a lot of that for you in Unified Measurement.
And by mixing the deterministic data that we still can get, and modeled measurement techniques like incrementality, and insights based on smart data governances, marketers can still optimize their ad spend effectively.
Deduplication becomes the key when platforms model conversions
AEM and Advanced Conversions and Ads Conversion Modeling and all the other variants of platform iOS mobile measurement modeling are important in 2025. Yes, they’re a little black boxy. Yes, it’s challenging to know exactly how the big mobile adtech players are modeling their results.
But we are seeing more insights as a result of modeling that the big platforms are doing on conversion reporting, even if they’re a response to programmatic platforms using probabilistic methods and aimed at recapturing some of the installs and conversions that SANs won attributions on in the past.
The challenge is that when you’re looking at Advanced AEM or Ads Conversion Modeling from Google or Advanced Conversions from Snap, how do you know that they’re not all claiming the same installs?
“ Is there actually one way or one source of truth around how you can actually look at your performance data now?” asks Beenen. “That goes diametrically against the reality on the ground when it comes to measuring iOS signals because there’s too many sources available … the nitty gritty really goes into how you are doing your deduplication.”
Deduplication becomes key, and that’s where additional data from ad partners, like Advanced SAN from TikTok, is helpful.
“ We’re supporting AEM on TikTok, we’re supporting AEM on Meta as well,” says Beenen. “And if you then total the AEM up, of course we see a jump in the actual campaign performance.”
But it’s also important to look at data in your MMP dashboard, like Singular’s, to look for assists, showing you when ads on competing networks are working together to generate a result. And it’s important to check Singular’s attribution decisions, which are based on all the measurement data we can access. Singular sees what each network claims, but then makes an attribution decision that deduplicates all those claims, providing your best view of real, actual impact.
Ultimately you do want to bolster your iOS mobile measurement with incrementality testing as well. When done right, incrementality testing gives you extremely reliable data on what different ad partners and campaigns actually achieve.
Because ultimately you want data from your ad partners — or at least what they’ll share in a privacy-safe way with Singular — but you don’t want them completely grading their own homework.
Creative measurement requires smart data governance
The challenge with all modeling is creative measurement and optimization, which you’d really like deterministic data on. One solution for that lies outside the platforms and networks themselves, and it’s data governance.
Seemingly unnecessary for smaller organizations, data governance is essential for all because campaigns and creatives easily reach into the hundreds, even on just 1 platform, even at small scale. For large organizations, we’ve seen 10s of thousands of creatives, which very quickly scales out of control if you try to manage it manually.
Data governance helps with iOS mobile measurement because when you manage creative names, campaign names, geo encoding, conversion outcomes, and any other metadata you want for optimization and reporting well, you can get high-quality first-party data on which creatives and CTAs and campaigns worked best, converted most frequently, and resulted in the highest ROI.
Even, to an extent, inside platform black boxes that might take your creative and mix it up in the blender to build thousands of ads from dozens of components. You’ll still know that a certain image and a certain string of text were involved in good — or bad — results.
Conversion APIs (CAPIs) are increasingly important in iOS mobile measurement
When you think CAPIs, you think retail typically: a person clicks an ad, goes to a website — usually a mobile website, but not always — and buys a product. A web SDK in that retail site informs the ad network, and a conversion gets claimed.
But it can be used in gaming as well, via a mobile SDK.
And it’s another useful signal, says GeekLab’s Jesse Lempiainen:
“ That’s something that we’ve utilized quite a bit, both web2app campaigns that we partner up with Meta officially, and helping mobile publishers and developers do those web2app campaigns,” he says. “ We’re kind of staking a step back and measuring everything on a creative level and, and then utilizing these web technologies essentially … so conversion API to send events back to networks.”
Success that GeekLab is seeing there includes:
- Lower cost
- More daily new users in App Store Connect
- Retention rate going up
But don’t expect it to be perfect: Lempiainen estimates the data accuracy level at around 70%.
Which brings up a good point in iOS mobile measurement for 2025: don’t expect perfect. And you don’t need perfect either. You do need good directional data that you can make quick decisions on … and that you can get.
In fact, that’s enough to keep the growth machine moving in the right direction.
(See what Jonathan Reich, CEO of Zedge, says about the perceived need for perfect data.)
More in the full podcast
Of course, there’s more in the full podcast episode. And don’t forget to check out all of the Hack Gaming Growth sessions as well.
Here’s what else you can expect:
- 00:00 Introduction to 2025 iOS Measurement
- 01:01 Current Challenges in iOS Mobile Measurement
- 03:50 Retargeting Strategies and Banner Ads
- 04:40 The Role of SKAdNetwork and AEM
- 05:47 Probabilistic Attribution and Industry Trends
- 09:06 Unified Measurement and Incrementality
- 19:24 Web to App Campaigns and Conversion APIs
- 24:36 Future of iOS Game Marketing in 2025
- 27:45 Conclusion and Final Thoughts
Get it wherever you get podcasts, or subscribe to our YouTube channel.
