Note:

We launched the Mobile Growth Lab where over 60 marketers, executives, product managers and developers signed up to break the shackles of ATT’s performance and measurement losses. While the premium and executive editions are closed for registrations, you can get access to the recorded versions of these sessions through our self-serve plan.

Check it out here: https://mobilegrowthlab.com/

ABOUT ROCKETSHIP HQ: Website | LinkedIn  | Twitter | YouTube


FULL TRANSCRIPT

If you’ve run SKAN campaigns, you will have likely seen ridiculously high CPAs(or low ROAS-es). We’ve seen campaigns go from a CPA of $40 or $50 pre-ATT to $700 or $1000+ post ATT.

Now, one reason for this is obviously Apple’s system of timers in SKAdNetwork. When a user installs, a 24 hour timer is started – and the first event is recorded that is completed by a user within a 24 hour time window. Subsequent events get recorded IF they occur within 24 hours of the preceding event – otherwise they are not recorded as coming from the campaigns.

As a result of this, many events(purchases, registrations, trials) that happen outside of the first 24 hours do not get recorded – resulting in high CPAs and low ROAS-es.

BUT that isn’t the only reason for insane CPAs and ROAS-es. There is another – and more significant, yet insidious one – this is the privacy threshold(which is also referred to as crowd anonymity in Apple’s WWDC session on SKAdNetwork 4.0). 

According to Apple’s WWDC 2022 session: “When the install count is low, we take extra steps to protect privacy by limiting the trackable information sent back. As the count scales up and the user’s uniqueness starts to blend into the crowd, we send more data back. Finally, as the count reaches the highest tier, we are able to send the most data back while still preserving privacy.”

What that means in practice is that if your campaign has very few installs per day, Apple believes that there is some leeway for you to reverse-engineer the user’s identity – and so Apple will censor conversion information for many of the users. Let’s say 2 users completed a purchase(and that has a conversion value of 5), then one of these users will trigger a postback with a conversion value of 5 – and another will trigger a postback without any conversion value at all.

So the lower your daily install count, the more the number of users whose conversions are censored – and the higher your CPA. According to a study done by Appsflyer, if you have less than 10 installs a day per campaign, close to 90% of your conversion values are obfuscated, which is why the CPAs can be scary.

Note that the privacy threshold determination happens at the time of install, and not at the time of a post-install event such as purchase or registration – so the privacy threshold is determined by the number of installs and not the number of post-install events. So it doesn’t matter how many purchases or registrations you get, privacy thresholds are determined by the number of installs per day.

The guidance we’ve gotten from platforms so far is that you’ll have minimal censoring of conversions if you have a minimum number of daily installs.

For Facebook this is 88 installs per day, Snapchat 75 installs per day, and TikTok 90 installs per day. Bear in mind, these install counts are per campaign. Why are these thresholds different for different sources? Because the way Facebook defines a SKAN campaign could well be different from the way Snapchat defines a SKAN campaign(for instance, Facebook might treat a Facebook adgroup as a SKAN campaign – and Snap might treat a Snap campaign as a SKAN campaign).

Now, all of this is in the current SKAN paradigm – SKAN 3.0 – where we have up to 100 SKAN campaigns(which can be defined by the platforms – Facebook, TikTok and Snap).

What happens in SKAN 4.0?

While we’re waiting on documentation still, what is known is that the 64 bit conversion value in SKAN 3.0 changes to a 2 part conversion value – a fine conversion value that is essentially same as the SKAN 3.0 conversion value, and a coarse conversion value that can take 3 values – say high, medium, low(which could be mapped to an early monetization or engagement signal – say trial, consumed content or canceled trial).

Again, according to Apple’s WWDC session: “at the low end of crowd anonymity, you will not receive a conversion value in your post back. At the medium level, you will receive the coarse conversion value. And at the highest level, you will receive the fine-grained conversion value in your postback.”

So: the mechanics of privacy thresholds(or crowd anonymity, depending on how you call it) will change – but the underlying principles won’t. What you will continue to see is that at low volumes a lot of your conversions will get obfuscated, and at high volumes, very few of your conversions will – and you’ll really want to consolidate campaigns as much as possible in order to maximize the conversion signal you receive.    

A REQUEST BEFORE YOU GO

I have a very important favor to ask, which as those of you who know me know I don’t do often. If you get any pleasure or inspiration from this episode, could you PLEASE leave a review on your favorite podcasting platform – be it iTunes, Overcast, Spotify or wherever you get your podcast fix. This podcast is very much a labor of love – and each episode takes many many hours to put together. When you write a review, it will not only be a great deal of encouragement to us, but it will also support getting the word out about the Mobile User Acquisition Show.

Constructive criticism and suggestions for improvement are welcome, whether on podcasting platforms – or by email to shamanth at rocketshiphq.com. We read all reviews & I want to make this podcast better.

Thank you – and I look forward to seeing you with the next episode!