💎 Growth Gems #116 - Data (AdAttributionKit) and Monetization
Fellow growth practitioners!
I’ve dug up more gems 💎 during my last shift in the growth mine. This week, I’m bringing you back insights on:
Mined from the depths, I hope these hard-dug insights from Felix Boudreau, Eric Seufert, and David Philippson will prove useful!
🥇 TOP GEM OF THE WEEK
SPONSORED INSIGHT & RESOURCE
RevenueCat won’t stop at being the leader in its category and producing some of the best content on its podcast and blog… it’s now also launching awards!
The Shippies are RevenueCat's first awards for outstanding achievements in subscription apps.
The awards are open to all subscription apps, regardless of whether they use RevenueCat.
It's a great opportunity to get your efforts recognized…and some sweet prizes.
The prizes? 150 digital billboards in San Francisco, a custom-designed Shippies trophy, and bragging rights.
I can’t wait to see the list of the best subscription apps in the industry!
Monetization: a case for hard paywalls
Gems from Felix Boudreau (VP of Growth & Marketing at Pok Pok) in Hard Paywalls: How New Apps Can Thrive in a Crowded Category at MAU Vegas
I’ve seen a lot of discussion about free trials vs. no free trials, which is a worthwhile consideration.
Much less is said about having a hard paywall: locking everything (or almost everything) behind it.
Felix makes a good case for it, and it might become a best practice for new subscription apps.
Measurement and reporting
💎 A lot of apps have more money than you, more content, more brand recognition, and a better presence on the App Store. Choosing the right kind of paywall for your app can help you stand out.
(02:25)
Felix describes soft vs. hard paywall this way:
Soft paywall = dismissable, and then you can explore (with some content locked out)
Hard paywall = non-dismissable, i.e., you enter a trial, or that’s it
💎 Pros and Cons of Soft paywalls:
Pros:
Bigger user base
Ad revenue
Drive network effect
More word of mouth
Better store ranking
Cons
Adds complexity
More customer support
More user = more costs
Worst conversion rates
(04:55)
💎 Pros and Cons of Hard paywall:
Pros:
Better initial conversion rates
Know your best users quickly
Less complexity (easier to understand LTV)
Simpler revenue model
Cons
Smaller user base size
Worst store ranking
Less word of mouth
Higher churn
(07:20)
One more advantage: it gives focus and puts a lot of emphasis on the critical onboarding part of the user experience.
💎 Quickly knowing who your best users are through a hard paywall is key for a startup because it allows you to set user interviews and find more of them.
(08:02)
This is similar to an insight from Jake Mor I shared in Growth Gems #101: “When you launch your app, it’s the worst it will ever be. So start by locking your entire app (with a free trial) to get the users that have enough of a pain point to be willing to pay for that app. Those are the early users you want to be building for.”
💎 A hard paywall is the right choice when you:
Have a small team with very limited bandwidth
Have some budget for paid UA (since organic won’t grow as much)
Can’t leverage soft paywall benefits efficiently (e.g., monetization with ads, limited network effect)
Having a hard paywall is a better mission fit
(09:20)
Below is PokPok’s paywall. Notice the CTA on the first screen…counter intuitive!
Other apps, like Kinedu, show a hard paywall after onboarding but then a closeable offer on app re-open.
💎 A hard paywall will likely lead to a higher install-to-trial conversion and a lower trial-to-paid conversion than a soft paywall.
(11:42)
Marcus made an interesting comment on this vs. “finding your best users”:
When you force users to start a trial (and inflate install-to-trial) while trial-to-paid suffers, it actually gets harder to find your best customers. As UA channels receive lower intent signal you end up finding more people that start a trial but don't convert.
Either option (hard/soft) will take 7 days to convert. If you want to quickly find your best customers, you'd need to eliminate the trial. Anyone who is ready to buy directly on d0 belongs in that group.
💎 Having a soft paywall means you have two flows, which decreases your focus as it gives many experimentation options. A hard paywall forces you to focus tests around the beginning of the funnel: number of screens during onboarding, prices, and subscription tiers.
(13:30)
My conclusion: a hard paywall is great for user research, focus, and improving your onboarding, but it might have negative consequences on UA in the long run.
So, it may make sense for a specific growth stage (early on).
Data: AdAttributionKit (AdKit)
Gems from Eric Seufert and David Philippson in Understanding AdAttributionKit on the Mobile Dev Memo podcast
The duo is back at it again, analyzing changes in the Apple ecosystem.
After their take on the Privacy Manifest and its implications, they discuss the introduction of AdAttributionKit (expected with iOS 18).
While there are no radical shifts vs. SKAN, it clarifies a few things and (maybe) gives a glimpse of the future.
I’ll do my part to spread the word: let’s just call it AdKit!
💎 StoreKit is the iOS kit for the App Store. This is AdAttributionKit (AdKit), which might hint that this is not designed to be a mobile-specific attribution network. In fact, there is AdKit for app and web.
(02:45) by Eric
Re-engagement *measurement*
💎 Re-engagement measurement, introduced in SKAN 5.0 at WWDC and now in AdAttributionKit (AdKit), is often misunderstood as re-engagement enabling. However, it does not allow to target a specific cohort of users. Instead:
It allows you to define an ad as a possible “re-engagement” ad (e.g., leading to an offer via a deep link “re-engagement” URL) if the app is already installed.
It allows you to measure/know which ads are seen by users that already have the app installed (via conversion-type = re-engagement)
It is still significant for big advertisers because you can measure new touch points. Example: Expedia or King can now know how many of their ads are served to users with the app installed.
(06:20) by David
Below are screenshots from the WWDC Meet AdAttributionKit video illustrating the re-engagement capabilities.
💎 You still have to show acquisition ads to people who don’t have the app installed because you can’t segment them out. There is no suppression list or re-engagement targeting, but at least you can measure.
(11:45) by Eric & David
In Apple’s example, the ad is the same for everyone. Then, users can be displayed a different in-app offer depending on whether the app is installed or not.
💎 Re-engagement measurement in AdKit is limited to 35 days post-download, even though the most valuable segments (more incremental) have usually lapsed a longer time ago.
(12:50) by David
Attribution becoming a USP for the Apple App Store?
💎 If Apple makes linking out of the App Store easier than currently, developers will have more reasons to launch on the App Store rather than an alternative store: they might get the best of both worlds. That is, if they also relax the fees they charge on off-platform fees and the Core Technology Fee (CTF).
(16:35) by Eric
In the EU, Apple relaxed rules on linking out but did not change the fees.
💎 AdAttributionKit (AdKit) becoming cross-channel (app and web) and (potentially) cross-device can be a unique selling point (USP) for developers to publish on the App Store rather than an alternative store. The integration of WebAdAttributionKit (“web” AdKit), previously known as PCM (Private Click Management), into the AdAttributionKit, enhances the App Store’s value proposition for developers.
(17:50) by David
The cross-device part is speculation, but some of these use cases could strongly benefit brands with a strong web and app presence. Apple has the OS and users’ Apple ID and could have true multi-touch attribution on the iOS ecosystem.
💎 Apple has to stop probabilistic attribution and fingerprinting; otherwise, AdKit is a waste of time. Apple also cares about perception, and consumers have said “no” to being tracked in ATT, yet they still get related ads.
(24:10) by David
💎 In AdAttributionKit (AdKit), only one view-through-attributed impression can be opened at a time: an impression starts, then stops, then another one begins.
(37:20) by David
David gave the examples of a rewarded video ad, followed by several interstitials. Currently, the interstitials have the same “view-through attribution” chances as the rewarded video. This won’t be the case anymore.
David considers this to be good for advertisers (more in line with customer behavior) as, currently, it’s possible to “game the system.” As a result, some ad networks expect view-through impressions to decrease and CPIs to increase.
Before I leave, here is a quote from outside of the mobile growth world that still applies very well (beware of proxies!):
“When a metric becomes a goal, it ceases to be a good metric” - Aaron Dignan (Founder of The Ready)
See you next time.
Stay curious!
⛏️ Sylvain
🔗 Sources: