iOS 15 and mobile marketing: your questions answered
The life of a mobile marketer is hard. iOS 14.5 just dropped, with massive changes, and now iOS 15 is here with more. On the other hand, constant change is kind of what we’re used to in tech, and it ensures that most of us still have jobs that haven’t been automated out of existence.
Plus, it keeps things fresh.
We brought four experts together for our recent iOS 15: The endless evolution of app marketing webinar, and they were really good. But we had so many questions we just couldn’t answer during the webinar. So … here you go.
First: our experts:
- Aykut Karaalioglu, CEO & Co-Founder, MobileAction
- Warren Woodward, Co-Founder & Chief Growth Officer, Upptic
- Yevgeny Peres, VP Growth, ironSource
- Gadi Eliashiv, CEO & Co-Founder, Singular
1. Spend trend away from Facebook and Instagram?
Q. What are you seeing in terms of spend on Facebook/Instagram ads since June? If it’s decreasing, where is the ad spend going? Thanks!
A. Yes, to some extent. We saw spend move to top tier ad networks that aren’t the biggest self-attributing networks, particularly those who put in a significant amount of work on SKAdNetwork.
We also saw spend move to Apple Search Ads.
It’s incredibly important, however, that you update your growth stack for SKAdNetwork to get an accurate picture of what’s working. Some mobile marketers have seen chaotic results in their growth and attribution data because some ad networks are finding ways to steal app install attributions via fingerprinting.
2. Google’s version of ATT coming?
Q. While Apple has been burning bridges and changing the mobile/ad ecosystem, Google’s business model is completely ad based. If you were to use your crystal ball, what do you see from Google’s more white-glove approach to ATT Lite?”
A. The key difference between Google and Apple is their revenue model. Google is a giant ad network, while Apple is a product and (increasingly) services company.
Google is going to do things that work for marketers, even while it is forced to build more privacy-safe advertising technology.
The question is whether the technology will be like FLoC, which leaned towards benefiting Google by providing knowledge to it over the interests of advertisers, or whether it will be more neutral. One technology that is extremely useful and is likely to have a bigger role in Android app attribution is the Google Play referrer, which Singular integrates. Google also wants to move away from last-click attribution in general, though it’s uncertain how that will be accomplished in mobile app install advertising campaigns.
3. Apple Search Ads ascendant
Q. I’m seeing ASA conversions skyrocket and all other partners, even the biggest ones, tank. Not to put on a tinfoil hat but is Apple grading their own homework here? Do privacy thresholds only apply to non-Apple partners?
A. First, see #1. Second, ASA is up significantly this year. Apple does have more data that it uses for Apple Search Ads campaigns than others do, and ASA also uses its own attribution engine, not SKAdNetwork. ASA also has a very generous 30-day attribution window.
So … all data you’re seeing is a version of reality. It’s a side of a coin. There are other sides (yes, plural).
The most likely thing to assume when all your data shows that the world is turning upside down is that because the industry is collectively changing how to score the game, we’re introducing unintended consequences. Also, that your ads are probably doing pretty much what they were doing a few months ago, but that the data you’re getting back has changed.
Somewhere in between ASA’s growth and the need to update growth stacks for new modes of attribution is like the truth.
4. Attribution data for different product pages in the App Store?
Q: Will we able to receive any attribution data for installs from different custom product page variants in the App store?
A. This is a fabulous idea and the answer is a definite maybe.
Here’s what we know so far from Apple: “Create additional versions of your product page with different promotional text, screenshots, and app previews to better showcase particular features or content within your app. Direct relevant audiences to a particular page using its unique URL and view performance in App Analytics.”
So you will be able to see the relative performance, and a specific ad will be able to send users to a specific version of your app’s product page.
How that data will translate to attribution data is not entirely clear, although obviously you could brute-force it by simply having separate campaigns for different product page variations.
5. App Store severance movement burning bridges?
Q. Do you feel there’s the chance that iOS may be burning bridges, in order to later launch their own controlled advertising as a counter strategy, should the App Store severance movement gain traction?
A. Full-on speculation mode: no. I don’t think that’s Apple strategy. That said, I do think Apple Search Ads wants to be the best place to do paid user acquisition. I’m not certain it can be, but it certainly is one good place.
6. Who canceled the privacy apocalypse?
Q. During a long time the industry called the IDFA deprecation a “Privacy Apocalypse.” Nevertheless, after iOS 14.5, when looking at aggregated data of the market (downloads, revenue) we don’t see a clear disruption on the trends or patterns. It would be great if some of the panelists can talk about this point.
A. Great point.
Just because we changed the measurement methodology of attributing mobile ads for app installs doesn’t mean actual consumer behavior changed or that the effectiveness of your ad campaigns changed. For those seeing wild swings: don’t panic, but do update your growth stack to fit the new realities.
Everyone else: getting the best data possible under the current data/privacy/measurement regime will optimize your chances for cost-effective growth.
7. Web to app measurement woes
Q. We all know the standard solutions for Web to app users on IOS 14.5. Now that Apple is blocking IP addresses and fingerprinting from trackers, are there any new solutions to track Web to app users?”
A. First off, the sky is not falling (yet). iOS 15’s Private Relay is optional beta software at the moment, only for those who are iCloud+ paying subscribers, and only for web-based traffic. That said, I personally think it will become the standard at some point and at some level of paid iCloud subscription, maybe/probably/possibly for in-app network activity as well as web-based.
Fingerprinting is verboten, according to how Singular sees Apple’s App Tracking Transparency rules.
Any possibility of doing that will get VPN’d away, most likely, or Apple will implement some additional technical solution to make iPhones present randomized signatures.
We are seeing marketers could adopt a partially web-based onboarding or registration flow, which would then give you attribution insight as well as multi-channel connection to your customers. See the webinar for a few more thoughts on that.
8. Massive iOS 15 Private Relay angst
Q. We know that Apple will eventually bring Private Relay to the furthest reaches of iOS (based on their history of privacy) but what can we do as advertisers/publishers to ensure our networks can still be viable and not suffer another ATT-style blow with LAT traffic?
A. Sorry about this in advance:
I’m guessing a Hunger Games kind of Haymitch quote like “embrace the probability of your imminent death, and know, in your heart, that there’s nothing I can do to save you” isn’t of much comfort here.
Look: the game has changed. Embrace SKAdNetwork for measurement. If you go web-to-app (because there’s some cheap inventory on web that’s worth trying) look for early onboarding opportunities as I mentioned above.
And, if all else fails, some of the best marketers on the planet still at least partially rely on the oldest trick in the book: actually asking people. (Who knew that worked when you want to know something?) Simply asking people how they found your app is useful and insightful, and worth adding to the texture of your existing attribution solutions.
9. An MMP to rule them all
Q: Which MMP is the best and why? 🙂
A: I swear this was an actual question.
10. I want my iOS postbacks now
Q: About the option to get SKAdNetwork conversions directly from Apple: do you already receive them that way? I believe this feature has been anounced but Apple has not actually started sending this data to advertisers.
A: It is live now.
11. Fingerprinting vs MMM vs SKAdNetwork
Q: Do you think Apple will force everyone to use SKAN soon? Many apps still use probabilistic or fingerprinting.
A: Since 75% of the mobile user acquisition universe is SANs, I think the answer is a clear yes right now. For the 25% of the universe that fingerprinting is even available for, I’ll say this: fraudsters are always on the alert for ways they can make you think the good things you’re getting are directly from them.
And I can tell you with certainty at least one major player is being pretty tricksy, as Gollum would say, about fingerprinting-based attribution.
Look: it’s against the rules, and eventually Apple will make technical changes to make it impossible. SKAdNetwork is here, it is deterministic, it can be predictive, and is pretty fraud-free when implemented well.
12. SKAdNetwork and subscription apps
Q: How are other marketers dealing with trial based subscription apps? Whilst tracking/modeling trials and installs is ok, what are the best practices for attributing post trial (sometimes 14-28 days!) subscriptions?
A: I’ve talked to a lot of marketers about this, and the answer is the same every time. It’s also one word:
You simply have to find early indicators of high-value users. Find something that predicts likely subscribers, or re-architect your onboarding or even your app itself if nothing is available. (No one said your job was easy.)
13. SKAdNetwork privacy thresholds
Q: What can advertisers who are finding themselves consistently caught by the privacy threshold do to try and get the most out of SKAdNetwork? To add, do you think Apple will provide more numerical clarity on the threshold in the future?
A: There’s only one real answer: concentrate your spend in fewer campaigns. You need at least 30 conversions per campaign to get decent results; more can be better.
And on your second question: nope. (Sorry, again.)
14. First party data for mobile marketing
Q: You mentioned getting more first-party data. How should we be looking to use that first-party data to improve monetization? Are there use cases about how to leverage it with ad networks we are working with?
A: First-party data is particularly important for platforms, major ad networks, and big publishers who can collect insights from tens or hundreds of millions of people gleaned from hundreds of different apps. That’s why you’re seeing so much consolidation in the mobile app ecosystem. (And it’s ongoing: Scopely just bought GSN Games from Sony.)
For your first question, assuming you’re a publisher with one game or app, first-party data will be important in your environment to know what turns people on and what turns them off.
This is valuable and important information for timing and placement of your ads and your in-app purchases.
It’s not something you leverage to any great extent elsewhere, however, which is why it’s trickier in terms of your second question: leveraging it with ad networks. Because unless you have permission to use that data as fodder for targeting insights with partner ad networks — in other words, unless you have a positive response to App Tracking Transparency — you simply can’t share it.
It’s first-party, so it’s yours. It’s not accessible for use on a third-party level.
15. Is this thing on? What’s going on here, anyway?
Q: Where will the answered questions be? I have to hop off soon for another meeting.
A: So glad you asked!
16. Anti-IDFA advice now, please!
Q: Any advice that the panel can provide for apps which have completely opted out of IDFA for iOS strategy would be appreciated!
A: I have good news and I have bad news.
The bad news (and, coincidentally, the good news) is that all the advice we can give you is pretty much the same as we can give anyone else who is presenting ATT and attempting to get IDFAs.
Since most app’s ATT success rate is basically at or around 20%, and since you need double opt-in on the advertiser app and publisher app for it to really work, and since that happens infrequently … we’re all in the same boat here.
Welcome to the club!
Go check the webinar for much much much more
We had some really smart people on the webinar, and they said some smart things. (It’s almost like those two things are connected.)
Go check out the webinar on-demand here, and play it in the background while you’re doing some work. When it hits an interesting point for you, tune in more intently. It’s pretty much a guarantee you’ll take away something of value for your work in the next month or so.