How to calculate ROAS in iOS 14 with SKAdNetwork
“First and foremost, I think we should recognize that there is still … IDFA, because users might opt-in,” says Singular VP of product Alon Nafta. “But if we go to SKAdNetwork — which let’s assume is the common case — it’s not like you don’t have anything. The SKAdNetwork framework does give you a few elements that if used correctly … advertisers can expect to get something sensible.”
Short version: yes.
Slightly longer version: there are some challenges.
Listen to the full interview on the Growth Masterminds podcast:
The elements Nafta is referring to include campaign ID, and publisher, plus inferred data like channel (via the publisher) and geo (via the IP). Cohorts, however, are a different story: not included by default. But they’re still possible: with the right partners and set-up, advertisers can get cohort data for iOS 14.
“That’s pretty much up to you,” Nafta says. “If you’re able to leverage the framework … that allows you to encode the date of install into the conversion value.”
Roughly, of course. This is iOS 14, and there won’t be a timestamp. The SKAdNetwork framework introduces essentially random delays, and theoretically if you wanted to keep updating conversion values, it’s possible that you would not see post-install conversion data for more than two months. But done correctly — and relatively quickly — it will give you a reasonable sense of cohort timing, Nafta says.
As long as you don’t expect too much.
“Real-time is pretty much gone,” he says.
The critical part is defining the most meaningful events that happen in your app and when they happen. For most mobile marketers, getting those in the first three days seems optimal, but you’ll have to test that assumption for your own app and your own business. A purchase would be great — of course — but any leading indicator of likelihood to purchase or otherwise convert is good enough.
It won’t be perfect. And, frankly, it won’t be as good from a marketer’s perspective as IDFA-based ROAS and cohorts. But it will work.
It’ll be good enough to optimize campaigns, Nafta says … and even creatives. (More on that later, in another podcast and blog post!)
Ultimately, that’s the best that marketers are going to get. Simply put: the world has changed, and without user-level and/or device-level data, Singular’s SKAdNetwork solution will provide what you need to grow intelligently.
Here’s the hard part: what if you want to optimize your marketing efforts for events that happen quite late in the user or customer journey? In other words, 30 days after install or even longer.
“I think that’s one of the questions that neither me nor others can answer in a perfect way,” Nafta says. “I would challenge you on some elements in the questions itself … you kind of assume that, ‘Hey, I can not change anything in the user behavior,’ but a 30-day trial, maybe I will challenge you a bit and say that … maybe now in this new reality, you have to change it to a seven day trial.”
In other words, your app experience and customer journey might need to change.
And marketers will have to work closely with designers, developers, and data scientists.
You don’t need a user transaction within that first seven days, but you’re likely going to see at least some user behavior that will be predictive for future activity, engagement, and conversion. App marketers who win will develop sophisticated models around those specific user behaviors that turn out to be good indicators of future monetization.
“The main thing to understand about SKAdNetwork is that conversion management is super important,” Nafta says. “Probably the most important thing. Sure, it’s still quite limited but we know that sophisticated advertisers are already planning to utilize smart techniques to try and leverage as much as they can from SKAdNetwork. And our goal with Singular SKAN is to have them do as little work as possible, and have Singular do most of the heavy lifting.”
Get more details on Singular SKAN here.
Or, request a demo.