UA’s Future: Google’s Privacy Sandbox vs. Apple’s SKAN

9 min read

In the world of mobile marketing, the development of privacy and attribution technologies by Google and Apple has taken center stage. Over the past several years, Google’s Privacy Sandbox and Apple’s SKAdNetwork (SKAN) have emerged as pivotal frameworks with distinct yet related objectives. Though both aim to accomplish some similar goals, the two sets of software and frameworks are actually very different.

Not surprisingly, therefore, what they mean for marketers is also very different, if related.

But because these changes are being driven by the makers of the two biggest operating systems on the planet, Android and iOS, learning to live in this emerging world is mission critical for people who want to grow their apps.

So what’s this brave new world going to look like?

Apple & Google: Building for privacy

Google’s Privacy Sandbox

Google’s software for privacy and marketing measurement is essentially Privacy Sandbox.

Designed for both the web (as third-party cookies phase out) and mobile (with GAID likely to be deprecated soon), Privacy Sandbox offers a comprehensive suite of adtech and attribution tools. It revolutionizes targeting, retargeting, audience building, measurement, and fundamentally enhances privacy.

Google’s tools include:

  1. Privacy Sandbox on Android
  • SDK Runtime
  • Topics API
  • Protected Audiences API
  • Attribution Reporting API

2. Privacy Sandbox on web

  • Pretty much the same as Android, minus SDK Runtime

3. IP Protection

Apple’s SKAN & privacy tools

Apple’s frameworks, regulations, and software for privacy and marketing measurement are not a 360-degree suite for adtech and marketing.

Rather, Apple’s efforts are a relatively disjointed collection of tools primarily focused on mobile, where Apple has much more significant investments. They don’t include functionality for targeting, retargeting, or audiences, but rather are primarily focused on privacy.

Of course, the part we hear the most about is SKAdNetwork (or SKAN), but that’s far from all of it.

Apple’s tools include:

  • Intelligent Tracking Prevention (ITP)
  • App Tracking Transparency (ATT)
  • Privacy Manifests
  • SKAdNetwork
  • Private Click Measurement (PCM)
  • Private Relay

Here’s what they look like, stacked up against each other and categorized both by their areas of focus and measurement impact:

Take a look at the potential role of ASO in privacy-friendly user acquisition

Why Google & Apple’s moves matter in mobile adtech

If the IAB was doing this, most of the industry would pay some degree of attention. If some medium or even larger adtech companies were doing this, some of the industry would at least check it out.

But Apple and Google are not small.

And together, they control the two biggest and most important operating systems on the planet, with somewhere around 5 billion users globally. Apple making iOS and Google making (most of) Android means that these ad mechanisms are baked into the operating systems of pretty much anyone who uses a smartphone, with some exceptions in China and a few other countries.

But it’s not just that Apple makes iOS and Google makes Android.

They also control the on-ramp to those operating systems: the gateway to billions of consumers in the App Store and Google Play, and therefore can enforce their dictates.

This may change over time due to legislation such as Europe’s Digital Markets Act.

Until then — and very likely after as well since billions won’t just instantly move to unproven new third-party app stores — Google and Apple will exert a huge amount of influence over the mobile ecosystem.

What’s the core difference between Google and Apple?

There are plenty of complexities and variances between Google’ adtech for privacy suite and Apple’s SKAdNetwork, but the core one is this: Google is an ad network, and built something to run advertising campaigns in a privacy-safe way, while Apple is a device and computing platform company, and built something to sort of enable advertising measurement while enhancing consumer privacy.

As an adtech exec once told me: Google built a 360-degree suite; Apple built some privacy tools.

If you look at it from first principles, when you boil down the privacy requirements of the digital marketing ecosystem, you need quite a few different tools to limit tracking. If you’re going to provide better digital privacy, you need:

  • Device ID obfuscation (IDFA on iOS, GAID on Android)
  • Device characteristics blurring (ITP from Apple, Privacy Sandbox from Google)
  • Device location masking (Private Relay on iOS, IP Protection on Android)
  • Privacy-safe marketing measurement (SKAdnetwork and Private Click Measurement from Apple, Privacy Sandbox from Google)

Google’s privacy sandboxes for both web and mobile are cohesive sets of technologies that provide much of what’s needed for an adtech ecosystem, but rebuilt for privacy:

  • Targeting (via Topics API and Protected Audiences API)
  • Retargeting (via Protected Audiences API)
  • Attribution (via Attribution Reporting API)

On the flip side, Apple has built 4 technologies for providing marketing measurement data to marketers in a privacy-safe way:

  • Private Relay (a VPN-like tool)
  • Intelligent Tracking Prevention (a third-party cookie blocker)
  • SKAdNetwork (an attribution framework for the mobile user acquisition industry)
  • Private Click Measurement (a conversion management framework for in-browser marketing)

In addition, Apple has built two policies that are enforced via its ownership of the App Store, the on-ramp to iOS app publication: App Tracking Transparency and Privacy Manifests. (Yes, they have technical implications, but they are more regulations than software or frameworks.)

In short:

  • Apple’s initiatives are more about mitigating adtech’s problematic capabilities; while Google’s are about reinventing the world within which adtech exists.
  • Apple focuses on mitigating adtech’s surveillance capabilities; whereas, Google focuses on rebuilding advertising within a privacy framework. This results in varying degrees of functionality and restrictions in their respective tools.

In addition, this is why adtech experts say Privacy Sandbox will break more than Apple’s SKAN, but ultimately be less disruptive. Time, however, will tell.

We’re currently testing Privacy Sandbox with multiple partners, and it’s not yet clear what degree of impairment marketers will see when the GAID goes the way of the dodo.

Marketers & privacy: What should you do?

Everything is getting more complex in mobile measurement.

For example, Google recently published a massive manifesto on modern marketing measurement, outlining what I call the measurement tripod:

  • Attribution: Micro-scale – the closest to actual engagement and conversions in both space and time.  It is about assigning credit for conversions.
  • Incrementality: Medium-scale and focused on determining the impact of marketing activity and spend. It is about justifying effort and expenditure.
  • Marketing/media mix modeling (MMM): Macro-scale, looking not just at organic marketing and advertising and spend, but also trends and local/global/vertical changes. It’s about looking at all the factors that impact revenue.

Historically, most mobile marketers have only focused on attribution.

Many are now starting to use or explore incrementality, thanks to new tools that make it easier than ever. And some are even looking at MMM, once purely the domain of big, wealthy, and ponderous companies that updated massive data-driven models quarterly at best but often annually or even on longer-term scales.

Here’s the future of measurement: multiple data points that help build up a probabilistic model of truth.

No one likes that, per se. We all like the perceived solid foundation of deterministic, granular attribution we thought we had in IDFA and still think we have in GAID (until Privacy Sandbox shoves that identifier off to stage left).

But let’s be honest – did we ever think that one single last-click measurement is truly and completely indicative of a real-world customer journey? (And yes, anyone who monetizes or supports monetization, whether via IAP, subscription, or viewing advertising is a customer.)

Of course not. Reality is more complicated. Causality is more nuanced.

It was good enough, but it wasn’t ever perfect. So, what we need now that most of its deterministic and granular nature is disappearing is multiple data points seamlessly mixed together to give us a better ground truth in marketing measurement.

Those multiple data points include:

  • Marketing spend data
  • Marketing delivery data (impressions, email views, deeplink opens)
  • Permitted granular measurement signals (IDFA, GAID, cookies)
  • Aggregated privacy-safe measurement signals (SKAN, Android Privacy Sandbox, ITP)
  • Revenue data (online and offline)
  • Customer data (engagement, events, CAC, LTV, cross-platform activity)
  • Ecosystem data (economy, weather, consumer behavior, seasonality)

That’s precisely what Singular is doing in Unified Measurement, a new way of building out a single source of truth that recognizes that no one measurement methodology — even IDFA or SKAN, even GAID or Privacy Sandbox — is sufficient to accurately represent what happens when your marketing, your product, people, the app stores, social media, word of mouth, accidents of fate, multiple touches, ASO, SEO, and everything else that goes into an actual go-to-market campaign collide in the real world.

We’re seeing good early results on iOS from marketers who have been using Unified Measurement for months, and we expect the same for Android in the near future.

Future-proofing yourself in Google and Apple’s world

You can’t control what Apple releases in SKAN 4 or 5 or 6. You can’t control what Google offers in Privacy Sandbox, or when it will deprecate the GAID or the third-party cookie.

What you can do is insulate yourself from changes in adtech and marketing measurement by spreading your bets across more measurement methodologies, plus use a measurement and analytics provider that captures all of that for you and summarizes it in a single report while providing full transparency to the underlying hard data.

Does that mean you don’t need to learn SKAN or Privacy Sandbox?

Absolutely not.

As challenging as they are, they are deterministic if not granular sources of attribution that provide important signals for measurement and optimization. They’re not the full solution, but they’re important parts of it.

What you do need to do is future-proof your marketing and attribution initiatives. That means not putting all of your eggs in one basket.

John Koetsier
by , VP Insights at Singular
John Koetsier is a journalist, analyst, and author. He’s a senior contributor at Forbes, VP Insights at Singular, and an angel investor.