Author: Gadi Eliashiv

• What is the future of marketing measurement?
• How will growth marketers measure impact in the age of privacy?
• What tools will brands need for both strategic and tactical marketing insight?

The age of privacy

We increasingly exist in the age of privacy, and privacy will increasingly define both how people engage with digital tools, and how marketers measure the impact of their efforts.

But the transition we’re making from deterministic granular data to a more nuanced aggregate reality is challenging. The word privacy has so many meanings and interpretations, and while almost all of them are overwhelmingly positive, in the realm of digital marketing, it also reminds people of uncertainty, complexity, confusion, and limitations. 

It’s the one topic that keeps coming up in every single conversation I have with mobile marketers. The one thing that constantly erodes measurement and attribution. The big caveat. The huge unknown.

Terms like GDPR, CCPA, third-party cookie deprecation, ATT, SKAN 1.0/2.0/3.0/4.0 and Android Privacy Sandbox are now at the core of every decision you make. Some exist thanks to a true concern for consumer privacy, while others are greatly shaped by the competing interests of today’s big tech platforms. 

But one thing is obvious: it’s a tricky landscape to navigate, and the job of growth marketing has only gotten more complex, not less.

For me personally and Singular in general, all this complexity highlights a very clear calling. Our job is to make things simple, and help companies achieve marketing greatness. Let’s be honest: both of those jobs are getting harder. 

But let me be 100% clear: they have not become impossible.

While it might be hard for those struggling in the trenches of SKAdNetwork to believe, I honestly see all of this change as a massive opportunity. And I’m fairly certain that if you are reading this article, you’ve probably seen that optimism across all our content … webinars, guides, Slack groups, workshops, and even in 1:1 consultations. You probably recognize that we spend a lot of time thinking about privacy.

My commitment to our community: we’ve got your back.

And today, two years into this privacy-obsessed era we live in, I wanted to take a moment and describe the vision I have for the future of measurement. But to do that, I have to start by describing what I see in the measurement space today, and how it shapes our POV towards the future.

A fairly divided market on the front line

While for now marketing on Android remains largely the same (and measurement is based on the soon-to-be-gone GAID identifier), Apple’s iOS is where things have changed the most.

In the past, most companies were pretty much on board with the IDFA-last-click model. Sure, we all ideally wanted MTA. We doubted last click. Sure, we wanted SANs like Google or Snap to share impression level data at the IDFA, so we could have proper MTA. Sure, we wanted the App Store to have a “referrer” mechanism similar to the Play Store. But for the most part – the market was aligned. You had an acceptable method of measurement.

But now – the market is heavily divided, again.

Division 1: iOS and SKAdNetwork

The first subject of division is SKAdNetwork (Apple’s privacy-preserving-yet-very-limited Attribution API).

If I had to cluster companies into two groups, I would define them as:

• Group 1 – yay! We got SKAdNetwork to work
• Group 2 – ughh! We can’t get SKAdNetwork to work

Let’s be 100% clear:

• Both groups contain super smart companies
• Both groups have companies of all verticals and sizes
• Both companies dislike SKAdNetwork (even if they got it to work!)

I’d even go as far to say that it’s quite popular to hate SKAdNetwork. I mean, it’s difficult, a black box, and easy to screw up. A lot of times you see numbers that don’t make sense (e.g. a $500 CPA), and your ability to make sense of it is limited. 

So why would you love it?

But here’s something that might surprise you: group 1 is bigger than you think. It’s just unpopular to be publicly positive about SKAN. And that makes sense, because even if you get SKAN to work, it’s still worse than what you had before in IDFA. Plus, if everybody else is struggling to make SKAdNetwork function when you can … it just might be a massive competitive advantage.

(Side note: if you’re in the group that can’t make SKAN work, talk to us. We have literally helped hundreds of companies become SKAdNetwork experts and they are getting amazing results.)

Division 2: probabilistic attribution (AKA fingerprinting)

Many companies (particularly ones with bigger brands to protect) have decided to rid themselves of any tracking-based probabilistic attribution that infringes on Apple’s guidelines. They are willingly accepting a potential competitive disadvantage to their business, because they want to play by the rules and stay safe.

On the flip side, everyone knows there are still companies that employ fingerprinting methods on a certain portion of their traffic. 

But let’s be real: it’s obvious that this portion is getting smaller and smaller. 

According to our data, SANs command 80%+ of the ad spend, and they are already 100% SKAN compliant and do not reveal data that can be used for tracking. The remaining 20% consists mostly of some fairly large and public ad networks, and these too operate with SKAN for the most part. 

So while some might be doing fingerprinting, they’re chasing an ever-shrinking part of the market. At risk of stating the obvious, this is not a winning long-term strategy.

Division 3: MMM and incrementality

In the wake of the IDFA deprecation, measurement technologies like Media Mix Modeling (MMM) and incrementality quickly became a hotly debated topic.

Media Mix Modeling is a data science-heavy process. It takes aggregate spending data, aggregate revenue data, other ecosystem parameters, and then outputs an estimated ROI by channel (or campaign, or more). The appeal is natural given its lack of dependency on IDFA, or any special access to platform data. Plus, it only needs aggregate spend and revenue data (something that’s very easy to get if you use a platform like Singular).

We all know the reality: MMM is not new, and has existed for many years now.

Predominantly MMM has been used to help Fortune 100 corporations answer very complicated media mix questions (think about connecting TV ad spend to body lotion sales at Walmart in Nebraska). What is new, however, is the new-found motivation to explore MMM, and see if it can be transformed from an almost archaic enterprise-only services-heavy solution, to a more modern light-weight SaaS product that can be made available to app developers suffering from the lack of insights provided by SKAN today and Privacy Sandbox on Android tomorrow.

Die hard fans of MMM believe this is the only way to look at your true ROI, and therefore measure the true incrementality of your media. Others simply say they are using a simplified MMM model for iOS applications simply because SKAN doesn’t work, and it can’t be relied on.

And while there’s definitely a growing number of fans, there are also plenty of critics: companies that tried MMM and found the numbers to be highly inaccurate (at least, not without a good deal of mostly manual ‘adjustment’). MMM can take a ton of investment to get right, and even its advocates have literally told me there’s still a lot more interest than actual adoption.

And incrementality?

It’s worth mentioning one more divide in regards to incrementality: some say the only gold standard for measuring incrementality is to run A/B tests against an audience, at the same time, and measure the conversion results of each cohort. The only problem is that it’s now extremely difficult if not impossible on iOS, and even in the Android world you can only accomplish it via intensive cooperation from your ad channels. So I’d be wary of tools promising you incrementality measurement through means of audience A/B testing in this era of privacy.

But let’s be clear: there is value in both media mix modeling and incrementality when adapted to the specific worlds and needs of mobile marketing. And, both can also be extremely valuable when used to measure and allocate spend in not just mobile and not just digital marketing channels but all marketing channels, including TV, connected TV, streaming media of all kinds, out-of-home, web, and even extremely traditional channels like print or flyers.

Hybrid measurement is the future for marketing

I think it’s obvious that we’re now at a point in time where the discussion shouldn’t be about which single measurement methodology we should use. Rather, it’s about the when, where, and how we use multiple measurement methodologies together.

That’s why I believe that the future of marketing measurement is what we’re choosing to call Hybrid Measurement. I could probably go on for hours about Hybrid Measurement, and truth be told – there’s still a lot more to learn and uncover as we put this to test with customers, but at its core, there are 3 key concepts we’ve settled on:

1. Unified data infrastructure
   
2. Multiple measurement methodologies
   
3. Reporting and insights serving multiple views and multiple purposes

Unified data infrastructure

You probably wouldn’t be shocked if I told you that at Singular we always believed that greatness in marketing involved bringing a lot of different signals into a single place. Since our inception in 2014, that has been the design principle of our product, and we successfully created the world’s best platform to deal with marketing data coming from literally thousands of different sources.

To properly build the hybrid measurement vision, there is a long list of critical data inputs:

• 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 & offline
• Customer data (engagement, CAC, LTV, cross-platform activity)
• Ecosystem data (economy, weather, consumer behavior, seasonality)

I can confidently say that our starting point in building this vision is very strong. Our unified data infrastructure is the best in the market, our existing offering of measurement methodologies and the sophistication with which we combine them is unparalleled, and our reporting and insights layers have been stress-tested and iteratively improved for almost a decade.

But that doesn’t mean we’re done. Far from it, we’re just beginning.

Multiple measurement methodologies

Instead of relying on a single view of performance (which already today is not really a possibility given the data fragmentation in iOS), there will be multiple views that employ multiple methodologies using all the data mentioned above, and serving different purposes:

• Views and methodologies based on permitted granular attribution data where available (Web, iOS, Android, 1st party data, PC, Console, CTV, Cross-Device data) that offer last touch or multi-touch attribution
  
• Views and methodologies based on aggregated privacy-safe measurement signals such as SKAdNetwork, Android’s PSA, Safari’s PCM
  
• Views and methodologies based on data science layers running on top of all the measurement signals available (such as our “SKAN Advanced Analytics”)
  
• Views and methodologies involving Media Mix Modeling which are useful for evaluating the high level ROI of channels and campaigns, determining incrementality, and informing decisions like budget allocation

And as you might have noticed, we’re increasingly adding layers of data science to our measurement methodologies. Nowadays it’s mostly focused around SKAdNetwork (with our SKAN Advanced Analytics product) and Media Mix Modeling. But in the future it will expand to additional privacy-safe technologies as they start capturing more adoption (e.g. Android Privacy Sandbox, Private Click Measurement, etc).

Reporting and insights serving multiple views and multiple purposes

The obvious question is: does hybrid measurement sum up into a single source of truth for all my marketing activity?

Right away? No. In the future? Yes. But it will take some time.

Off the bat, these data sets were not designed to fit with one other. Imagine taking permitted granular attribution data (e.g. IDFA), combining with aggregated privacy-safe measurement data (e.g. SKAdNetwork), and marrying it to statistical outputs from an MMM model. You simply have no clear way of identifying the overlap between these data points. One is device level, the second is an aggregated cohort level which is still somewhat deterministic, and the last is a complete statistical model looking at correlations between spend, revenue, and additional factors.

Therefore, merging them is extremely challenging to do today without a lot of hand waving. So far, any initial attempts we’ve seen to do so, even for a limited subset of this data (like merging IDFA and SKAdNetwork data) are overly simplistic and quite inaccurate.

As an industry, we’re still at a relatively early stage of proper SKAdNetwork adoption, and that means we’re a long way from having every app developer run their own MMM model successfully. Essentially there hasn’t been a platform up until now that has taken on the challenge of supporting all these multiple views in such a comprehensive way, so there are no real past experiences to learn from.

But, at the end of the day it’s very clear to us that marketers want and need a way to make sense of these numbers in unison, and that will be one of the goals for our data science, engineering, and product teams for the months and years to come.

When will Hybrid Measurement be released?

This will be a gradual process, and I envision three phases:

1. Phase 1: Enable access to more methodologies through our common data infrastructure and data science layers (AKA our Hybrid Measurement platform). The focus here would be on actual adoption as opposed to just hype.
   
2. Phase 2: Reimagine reporting visualizations for multiple methodologies. This is already being surfaced and worked on today, and will become even more important when more methodologies become available.
   
3. Phase 3: Develop ways to combine results from multiple views and methodologies into a single insight engine based on all available data sources. 

These phases will not be completely sequential, but it’s clear that learnings from phase 1 will greatly impact phases 2 and 3.

Final words

We are EXTREMELY excited about this vision.

Due to skill or luck (probably both :)) we have built the perfect platform to enable us to carry out this vision. It has always been the natural expansion path for our platform, and the industry as a whole, and privacy was the perfect catalyst to make it happen.

It’s beautiful to see how data science is becoming embedded in everything we do nowadays, and how things are getting more sophisticated. On the flipside, we must remain committed to simplicity and accessibility which is what the next generation of measurement solutions must provide.

Stay tuned as we’ll be sharing more about this over the next weeks and months.
And until then, you can find us in all the usual spots: slack, webinars and of course our blog.

Google has been working on their Privacy Sandbox initiative for a while, mostly focused on the web. Yesterday Google also announced their plans to expand the project to the Android platform. While the web sandbox is trying to remove third-party cookies (mostly because of outside pressure), the Android Privacy Sandbox project’s goal would be to deprecate the GAID.

Doing so would impact a lot of mobile marketing systems from measurement, targeting, retargeting, and more.

The good news is that Google’s proposal for measurement – as it stands right now – seems quite good. In fact, it’s better than good. It’s surprisingly thoughtful, flexible, and proves that you can get rid of GAID without wreaking havoc on the app ecosystem.

It might not be very surprising given Google’s deep roots and foundations in the advertising ecosystem, their inherent understanding of the space, and their desire to leave it unbroken. It’s probably also a lot of battle scars from trying to do similar things on the Web, that led them to a pretty great first draft.

Google says the expected timeline for the project is 2+ years – and we suspect that what we’re seeing today will keep on changing in that period. Nonetheless, I want to applaud Google for putting these ideas forth as proposals, ahead of time, and seeking feedback! We didn’t get a chance to do that with SKAdNetwork, and I think it’s such a shame …

The Android Privacy Sandbox itself has a few topics that address targeting and retargeting, but this article is going to focus mostly on the attribution stuff … because well … it’s a thing we care about a lot here at Singular 🙂

But first – a caveat:

This post is based on my initial interpretation of what is a fairly early and complex document released by Google. If you think I got a detail wrong, please tell me about it!

 

The TL;DR: Attribution with the Android Privacy Sandbox

The main idea here is that Google will eventually deprecate GAID, but still wants advertisers to be able to run user acquisition effectively. As the biggest ad network on the planet, Google knows measurement is key to that.

To do that, Google will create an on-device service that will be built-in to the Android OS. They call that service the “Attribution Reporting API,” and this service will have a few tasks:

• Store views and clicks reported by ad networks (Google calls these “sources”)
• Store conversion events like installs, purchases, signups reported by Apps/Singular (Google calls these “triggers”)
• Match reported conversion events with reported views/clicks that are stored on the device
• Send the data out to networks/measurement partners/advertisers in two forms:
  • Event-level reports: reports that contain super detailed upper-funnel breakdowns (campaign, sub-campaign, creative, down to the click_id itself) paired with super limited lower-funnel data (1 to 3 bit conversion value)
  • Aggregatable reports – a more balanced report that contain selected upper funnel breakdowns (GEO, campaign, etc) data with more detailed lower-funnel data (revenue, number of purchases, etc)

 

Event-level reports kinda sound like SKAdNetwork. Are they similar?

At first read – it seems way better than SKAdNetwok.

The “event-level reports” will provide a super granular breakdown of your upper-funnel data (think campaign, sub-campaign, creative, down to the click_id itself), paired with fairly limited conversion values (3 bit conversion values – vs SKAN’s 6 bits).

It does seem, however, that it will be possible to send up to 3 conversion events at separate times that will each be attributed to the view or click. So probably the first conversion event will be used for the install, while the subsequent 2 conversion events will be for meaningful KPIs that can happen later in the user’s lifecycle. That’s huge … and was one of the main features we wanted from SKAdNetwork.

Here’s a summary of event-level reports vs. SKAN:

 

 

What kind of reports can we build with event-level reports?

To make it simpler to understand, here’s an example of what a set of event-level postbacks could look like:

 

 

The “Source event ID” is defined by the network, and is essentially the click/view ID defined by the network. That means that this has a 1:1 mapping to all the data marketers care about, such as Campaign Name, Creative Name, Country, OS, Audience, etc.

The “trigger type” is defined by the advertiser/MMP, and is very similar to SKAN conversion values. It does have fewer bits than SKAN, but you can send up to 3 triggers at different times … so good news for cohort reporting!

Here’s a super simplistic example for how an advertiser/MMP can define “trigger type:”

• 001 = install
• 010 = level completed
• 101 = 7 day revenue > $10

Putting all this together, the result in a Singular dashboard could look like this:

 

 

So while the “trigger type” (conversion value) is more limited than SKAN, you get incredible granularity on the upper funnel data, and you get the possibility of sending multiple conversion events … and that by itself already provides a more powerful report than what we’re currently getting from SKAdNetwork.

And you know what – there’s more. 

Check out the next section about “Aggregatable Reports.”

 

What kind of reports can we build with the “aggregatable” data?

This area is particularly exciting, and it has a pretty clever implementation as proposed by Google, but there are also some vague details at this point.

 

 

The general concept is as follows:

1. When an ad network reports clicks and views to the Android Attribution API, they will include “aggregation keys” that will represent certain breakdowns. These keys can be up to 128 bits, which is super useful (128 bits is a TON). For example, imagine that ad network “MobCool” always uses 16 bits for their Campaign ID, and 16 bits for their Creative ID. And let’s say unimaginatively their Campaign ID is 1 and Creative ID is 2. To create the breakdown, they convert the Campaign ID and Creative ID to bits, and concatenate them. And this would result in a 32 bit string:
   0000 0000 0000 0001 0000 0000 0000 0010
2. Sometime later, when the app is installed, the advertiser/MMP will report conversion events from the app to the Android Attribution API. These events can also append their own breakdowns into the aggregation key. For example, if the advertiser wants to further break down their marketing reports by device model, they could build a key with 10 bits (I chose 10 arbitrarily – they can use 20 bits if they want), and it will look like this:
3. The bits from the ad network and the bits from the advertiser/MMP would get appended together.  So for a user that came from Campaign 1, saw Creative 1, and uses Acer Liquid A1, the breakdown key would be: 0000 0000 0000 0001 0000 0000 0000 0010 0000 0010 And this is only 40 bits. We have up to 128 bits … so imagine the useful breakdowns possible with this! (Note: we’re not sure yet if there’s a limit on how many different breakdowns you can have. 128 bits is very large.)
4. When reporting the conversion event, the advertiser/MMP must also report a numerical value for the conversion event itself. This can be a simple counter for things like installs (e.g. 1), or any number for things like revenue (e.g. $100). This value is what gets summed together, when the aggregation happens (see more about that later). (Note: there are limits on the values you can pass here. This is part of the differential privacy mechanism Google employs).
5. Every time there’s a successful match between a conversion event (“trigger”) and a view/click (“source”), Android’s on-device service will store this information, together with the aggregation key that was concatenated together. It will then send this user-level data in an encrypted form to the customer’s ad tech platform (MMP, ad network).
6. Now the customer’s adtech platforms (MMP, ad network, etc) have a lot of these “encrypted aggregatable user-level data” records. In order to aggregate them, Google came up with a cool idea. There will be a separate network server called an “Aggregation Service” that will simultaneously decrypt and aggregate the user-level data, based on these predefined aggregation keys. The end result from the above example could look like this:
7. And this is obviously simplistic. A proper implementation would add a ton more dimensionality, and at Singular this table would also be connected with various data sources to arrive at real ROI. Google is allowing the aggregation key to be as long as 128 bits – which means there’s a ton of room for adding meaningful breakdowns, and the resulting reports could be really good.

Plus, on the aggregation service: 

This “Aggregation Service” is a program built and signed by Google, but will be running on other companies’ servers in what’s called a “Trusted Execution Environment.” That means that if Singular is running an “Aggregation Service” we can’t manipulate how it works, and it’ll work how Google intended.

 

How will the attribution waterfall logic work?

We think we understand what’s going on, but as we said at the top of this article – there’s a lot of complexity and some unclear direction in the developer documentation as it currently exists. That will probably get cleared up as the project progresses, but not everything is 100% set in stone or fully explained.

Here’s what we think is happening:

1. Every ad network can report their views and/or clicks (which Google calls “sources”) of its ads.
2. Networks are also allowed to assign priorities for each of their touchpoints (e.g. most networks will probably agree that clicks should have a higher priority than impressions).
3. When a conversion event (what Google calls a “trigger”) is reported by the installed app, Android will try to match that conversion event with all the relevant views and clicks, and choose the one with the highest priority.
4. Apparently, this logic is done completely separately for each network… which seems to imply that if two or more networks had clicks and/or views (something that happens in real life) both of them will receive attribution postbacks (event-level and aggregated reports).

Clearly item #4 above raises the question of deduplication across ad networks. We can’t have all the networks winning (and getting paid) as this would skew the marketer’s view quite considerably by double counting.

The documentation specifically addresses third party measurement use-cases by enabling redirects on the view/click events. So this is how we think companies like Singular could achieve deduplication across networks for their customers:

1. When Ad Networks report clicks and views, they would use the “Attribution-Reporting-Redirects” header in the reply (see function registerAttributionSource) which will include’s Singular endpoint.
2. The API will then reach out to Singular’s servers, and we would register the exact same views and clicks based on our customer’s waterfall and prioritization choice (e.g. clicks are more important than views, the attribution window should be X, etc …).
3. Once a conversion event is reported by our SDK in the advertiser’s app, the Android API will match it against all views and clicks that we reported, and choose the relevant touchpoint based on the prioritization we supplied – thus achieving deduplication across ad networks.

If this flow is implemented in the way we proposed, this would also have the added benefit of providing some MTA reporting!

 

Open questions: so much more to learn

There’s a lot we have yet to dig into, and plenty that will need clarification from Google or the ecosystem as we progress. Here are a few open questions for me:

1. How will we do fraud detection, management, and mitigation?
   For now, it’s unclear what prevents companies from claiming that they got a response from Google. We didn’t see any mentions of cryptographic signatures, as we have in SKAdNetwork.Also, how does this API prevent malicious parties from registering endless clicks and impressions whenever they’d like? (In other words, click spamming.)We did notice the registerAttributionSource function expects an “InputEvent” to register clicks, but not sure how it verifies this was an Ad Click, and not just any InputEvent. Also – for views, there is no such need for an event.
2. Why URLs versus parameters?
   Why do the registerAttributionSource and triggerAttribution methods always require receiving a URL versus receiving the parameters directly? (We assume it’s some protection as the URL resolution will be done in another process, but can’t be sure.)
3. How will the industry coordinate on aggregation keys?In the section above, we explained how aggregation keys are a combination of values from the ad networks (that define part of the key on views/clicks), and advertisers/MMPs (that define the other part of the key on conversion events). The API enables the advertiser/MMP to overwrite all breakdowns, and caution must be taken so that we don’t overwrite something the ad network needed by accident. This creates some complexity, and perhaps something Google could improve in the API, so that it won’t require coordination between a ton of companies. (64 bits for networks, 64 bits for advertisers/MMPs? :))
4. How will Google Ads use this functionality?
   Apple has a different process for Apple Search Ads than other ad networks have to follow. Will Google Ads follow this new Privacy Sandbox for Android just like every other ad network? Early indications are yes, but there are still some questions.
5. How can mobile app deferred deep linking functionality be supported?
   Deferred deep linking has traditionally been supported with server-side attribution processing, Google’s current install referrer framework, and even some on-device methods.  However they all could also be used to circumvent the privacy goals of the Privacy Sandbox. We would like to see a deferred deep linking solution built into the APIs to help advertisers deliver the streamlined user-experience for their new users.

 

Summing up

It’s early days. Google literally just announced this initiative, and clearly there’s going to be a lot of iteration. We’re excited that Google is looking for feedback, and we’re excited to collaborate with them on this initiative. The early documents show a much healthier approach that can offer dramatically better reporting than SKAdNetwork, which is very encouraging. Who knows, maybe there’s even a few things here that Apple might include in SKAN …

One thing is sure, you can count on us to keep covering this, and iterating on it with Google and our customers.

If you’re looking to discuss this further, we highly encourage you join our professional Slack group on all matters of mobile attribution privacy.

Facebook’s Advanced Mobile Measurement program is officially over.

This means that the granular device-level data you used to receive is going away in favor of more privacy-safe aggregated reporting. Facebook has deprecated Advanced Mobile Measurement (AMM), and now advertisers will only be offered aggregate install measurement from Facebook, not granular data.

Privacy: win.
Marketing measurement: loss.

But there’s good news for marketers as well as users here.

You can still get device-level data for your Android app installs. And, little secret: it’s actually a net positive in a number of different ways. At least if you use an MMP that supports Facebook’s new Google Play​​ Install Referrer solution out of the box right now.
(Which, thankfully, Singular does.)

I wanted to take a moment to share what’s happening, my perspective on it, who this impacts, and how we will continue to be your partner in growth.

And, I wanted to give you all the details on how Facebook’s new Google Play Install Referrer measurement solution solves many of the issues arising from Facebook AMM getting less granular.

What is changing: granular measurement access

As we’ve seen very clearly, 2021 is the year of increased privacy on mobile. The most recent big event here was the launch of iOS 14.5 and the ongoing transition to SKAdNetwork as the primary source of attribution truth for paid iOS app marketing.

As of October 29th, 2021, Facebook is making a change as well: removing device-level data from advertiser view.

In other words, Facebook install measurement information will only be available in aggregated reporting. That means no IDFA on iOS or partial user-level visibility on Android. It means more end-user privacy, of course. And it also means changes for how some mobile marketers run user acquisition and their internal measurement, BI, analytics and machine learning.

Importantly, the data MMPs like Singular receive from Facebook will remain unchanged. We’ll continue to be able to provide data on conversions at an aggregated level. (See below for more).
And there’s even more good news: Google Play Install Referrer access.

Goodbye AMM, hello Google Play Install Referrer

Instead of providing device-level data in the AMM program, Facebook has decided to introduce a Google Play Install Referrer measurement solution.
That works pretty much like an HTTP referrer would on the web:

•  A user clicks an ad
• They go to the Play Store and install the app
• Once they open the app, Singular can see the click metadata and assign it to a Facebook campaign
• The result is that you get independent device-level performance data on your campaigns

As I mention in this post on the Google Referrer change, there’s some very significant upside here which takes away some of the sting of Facebook AMM going away:

The first obvious win is that advertisers can get back much of the Facebook attribution data that was available to them via the AMM program … this means that a lot of disruption to BI/internal analytics systems can be avoided …
This also opens the door for longer cohorts.
Facebook device-level attributions must be deleted after 180 days …. Google does not provide any clear retention requirements for Install Referrer data, which means we’ll be able to offer longer cohorts (e.g. 365-day) for app users.

What that means for you essentially is more granular data on campaign performance for longer periods, providing improved insight for marketing optimization. Existing user-level postbacks and ETL destinations will automatically contain this data once you configure it in your Singular dashboard, and we’ll maintain the Facebook self-attributing integration so it’s available to compare and contrast.

How you can still do granular analysis

It would be foolish to say that Facebook’s AMM deprecation isn’t a big deal, and doesn’t present some challenges for marketers. But there are two important things to keep in mind when you still want to do granular marketing measurement analysis.

First, as we’ve already mentioned, the Google Play Install Referrer data offers some of the data back that you’ll lose. It’s Android only, sure, but iOS is already gone anyway due to SKAN.

Second, Facebook is offering access to insights from granular data via MMPs, even if not the granular data itself.

As an MMP, Singular still has access to device-level parameters for app install campaigns. Per Facebook policy, the device-level data cannot be shared, but Singular can still process and combine it with your other data sets, at which point we can share these aggregate insights.

As an example, we could run something like user-level LTV predictions, then share aggregated insights back to you at the campaign level.

MMPs like Singular are certified to handle data privately and in accordance with appropriate rules: national, international, and program-based (like Facebook’s MMP program). We are literally audited on our ability to access, ingest, store, and transform this data safely while abiding by the rules. Therefore, as I mentioned earlier, Singular will still be able to function as a trusted partner: by our customers on the one hand, and by the world’s marketing platforms on the other hand. The data we will receive from Facebook will remain unchanged.

That’s where the innovation comes in.

Singular has always been the world’s best MMP at ingesting, processing, and combining data from multiple sources. Now that’s simply going to be extended.

You’ll of course always have access to your first-party post-install data: what new users are doing. Singular will have device-level parameters for who installed your app. While combining the two can’t be done in your servers, it can be done in Singular servers. We can’t share the device-level data, but we can share measurement insights derived — at least partially — from it.

That changes the game.

All of a sudden, the MMP you choose must be world-class at ingesting more of your first-party data to combine with the available-but-not-shareable device level marketing results data. And then MMPs need to allow you to build on-platform measurement models and engines that provide predictive insights.

In other words, what matters now is inputs and outputs, and what you do with the outputs to feed your acquisition engines. This is a big part of what we call next-generation marketing measurement.

Now, more than ever before, mobile marketers with scale need an MMP with world-class measurement capabilities as a trusted partner. That means being able to bring in your CRM data, your models, your post-install insights, combining that with device-level data in a private, sandboxed container, and exporting the resulting aggregated insights to you. That likely looks like brands delegating more of what was previously internal measurement to Singular, but the results will be similar.

For instance, LTV prediction has generally been run at the user level, but the results are typically applied at the campaign level. Singular can connect customers’ user-level LTV predictions with the user-level install data and provide aggregated campaign-level insights based on that. This offers high-performance predictions while avoiding user-level data leakage, supporting advanced insights.

This is clearly a change for some, and it may cause some challenges. The good news is that data management and processing is literally a core competency at Singular. We know how to do this. It has been in our DNA since day one. And we’re committed to being your trustworthy measurement partner through this and future changes.

More questions about AMM deprecation?

You’re probably wondering what the impact is and what to do about it. Here are a few thoughts.

1. Facebook campaigns won’t be impacted
The first and most important point: your Facebook campaigns will not be impacted. This won’t harm performance for Facebook campaigns. This may sound trivial but it’s important to highlight: everyone is concerned with user acquisition budgets, uncertainties, and performance.

2. This is primarily an Android platform change
As you know, post iOS 14.5, device-level data is essentially over for mobile marketers on iPhone and iPad.

While it’s true that some double opt-in users who say yes to Apple’s App Tracking Transparency (ATT) pop up in both Facebook and your app could result in IDFA visibility, the reality is that this is a small percentage of activity. Even so, some have definitely done that, so this change will impact the use of methods leveraging those opted-in users as a survey sample size for the entire population of their app installers. And those will now only be possible via your MMP.

Besides that, however, this is primarily about GAID access.
And it’s a big deal because Facebook is clearly one of the two most important companies on the planet for mobile user acquisition campaigns. Most iOS device-level data is gone, and this change takes a significant bite out of the data mobile marketers thought they still had on Android.

3. But … this is not really an ecosystem change in the same way iOS 14.5 was
iOS 14.5 demanded changes from ad networks, SANs, measurement partners, mobile marketers themselves, and more. There’s almost no part of the mobile marketing environment that wasn’t impacted in some way, and specific ones — like retargeting and granular reporting — were almost terminated.

But with this new change, ad networks are not impacted, Facebook itself has very minimal change to deal with, and MMPs don’t have a heavy lift — at least initially — to support the change.

4. And … this change doesn’t hit everyone equally
If, for instance, you mainly rely on your MMP for reporting and you use data from the Singular dashboard or ingest aggregated data to your internal BI stack, there’s essentially no change. All the insights you depend on to optimize remain, and only having access to aggregated reporting won’t really affect your decision-making capability.

However, for very advanced publishers with big data science teams and sophisticated machine learning models that rely on attributed device-level data, deprecating AMM is a change that requires a paradigm shift.

If your internal measurement dashboards are built on AMM data, they may not work. You will have to look at the Google Play Install Referrer data as a means of getting some of this back.

5. Important: ultimately, we believe this is not just a Facebook event
In all likelihood, we will see other platforms follow the same track.
Today’s privacy implications are the same for every major platform, and while Facebook is leading the change, I fully expect that additional key players will follow suit. Multiple platforms are likely going to change how user-level data is reported and under what conditions advertisers can access it, and so the industry might as well get ready now.

Talk to us

Change can be painful. Please reach out to us to talk through what this will look like, what it will impact, and how you need support from Singular.

We’re always happy to chat, and I’m personally more than happy to take some time and walk through the implications for you.

Earlier this year I shared a clip from Inception that illustrates the change we’re experiencing in the mobile marketing ecosystem. Clearly we’re still in that reality.

 

 

But sometimes the changes are positive.

 

What’s happening: Facebook launching Install Referrer for Android app campaigns

A few weeks ago, Facebook announced that they are deprecating their Advanced Mobile Measurement (AMM) program within their Mobile Measurement Partner (MMP) integrations at the end of October 2021, such that advertisers will only get aggregate install measurement for Facebook app install campaigns from their MMPs.

While the impact on iOS (post iOS 14.5) is expected to be minimal since AMM data is not available today for iOS 14.5+ opted-out events (not all devices have IDFA anymore and are therefore not eligible for device-level attributions in AMM) — the impact on Android is expected to be greater. (We had a great webinar discussing the impact here).

Recently Facebook announced that following successful testing, they intend to leverage Google Play Store’s Install Referrer to bring back some of this data for Android.

How it will work technically:

• You run an Android app ad on Facebook
• A user clicks on the ad and is directed to the Google Play Store
• Facebook appends secured campaign metadata to the click that gets passed to Google
• If the user installs the app within the Google Play Store, the Install Referrer will pass that secured campaign metadata to the advertiser
• Once the app is installed and opened, Singular can retrieve the secured metadata for clients and use it to retrieve information about which campaign directly led to the app install

(Note: this isn’t changing any attribution data flows, since the data Facebook provides MMPs will remain unchanged with the upcoming deprecation of AMM. Also note that as Facebook is still testing the use of Install Referrer, further clarity and more details will be provided as they move to a general roll out.)

To provide some insight, here is a quick FAQ style article about this, based on what we know so far.

 

Is this data set exactly like the old data set?

Sort of, but not exactly.

Clicks that did not lead directly to an app install will still be counted as attributions by Facebook and MMPs, but won’t be available in this data set.

 

What’s great about this change?

The first obvious win is that advertisers can get back much of the Facebook attribution data that is available to them via the AMM program which is being removed at the end of October 2021. This means that a lot of disruption to BI/internal analytics systems will be avoided … at least on Android.

This also opens the door for longer cohorts.

Facebook device-level attributions must be deleted after 180 days (we actually get audited for this). Google does not provide any clear retention requirements for Install Referrer data, which means we’ll be able to offer longer cohorts (e.g. 365-day) for app users.

 

Speculation: Why did FB make this change?

When we at Singular think about Google’s eventual privacy changes on Android, we predict that Google will:

• Deprecate the GAID, giving a huge boost to privacy by killing device graphs and data collectives and retargeting
• Keep the Install Play Referrer mechanism, enabling advertising and measurement without trading persistent device identifiers

In my opinion, Facebook is thinking the same – and they are preparing for the eventual GAID deprecation.

Another point of view that we discussed internally at Singular is that Facebook is basically adhering to the privacy-preserving mechanisms each platform offers:

• ATT on iOS 14.5+
• Google Play Referral on Android

 

How will that work with my Singular/MMP setup?

Since we’re already collecting Google Play Referral data (it’s a widely used mechanism with other ad networks), we’ll pass back the information to advertisers with the same data pipelines we had before the AMM change. This is great news for advertisers since there will be no additional work required to receive the same data sets.

This probably won’t change how we do attribution however, since MMPs like Singular will continue to have the same data access from Facebook, which provides a more accurate and complete representation of the Facebook attribution claims, including:

• Impressions and impression based attributions
• All clicks, and installs derived from these clicks based on the attribution windows

The other change to look out for is the possible capability to measure longer cohorts and true LTV on Android users acquired from Facebook. We’ll have to intelligently combine the full data set we get from the MMP endpoints – which has certain retention rules – with the data provided on the Referral Mechanism, which might have different retention rules.

(Since this is still TBD, I’ll refrain from making any promises just yet!)

 

Summary

While we live in very turbulent times, and there’s a lot of changes, I believe this is a very positive change by Facebook.

At this point I’d like to extend my promise to our wonderful customers and the industry as a whole that we’ll continue to help being your measurement partner across all your ad networks and marketing channels, aggregating insights from multiple different sources and multiple different datasets to provide you with a single source of truth.

That includes spend data, self-reported platform performance data, deterministic device-level attributed data, deterministic aggregated data like that from SKAdNetwork, probabilistic aggregated data for cohorts and media mixes and incrementality, referrer data, and more.

Referring back to the GIF I posted at the beginning of this article, I’d like to think that we’re your gravity boots. We’ll strive to provide stability, continuity, and education on industry changes like this one … so you can quickly catch up and stay ahead!

Until next time!

– Gadi

ANGUILLA, B.W.I., June 28, 2021 – Kidoz Inc. (TSXV:KIDZ) (the “Company”), mobile AdTech developer and owner of the market leading KIDOZ Contextual Ad Network (www.kidoz.net) and the Kidoz Publisher SDK, is pleased to announce a new partnership with Singular (www.singular.net), a leading marketing analytics and attribution platform, to measure user acquisition and unify marketing data in the kid-safe mobile market.

The challenge is to enable app developers and marketers to spend intelligently on campaigns to acquire new installs for their apps in a fully compliant and kid-safe way.  Singular and Kidoz, both certified by PRIVO and compliant with all COPPA and GDPR regulations, have partnered to enable Singular’s privacy-compliant child-safe attribution product with Kidoz’s Contextual Ad Network.  Employing Singular’s COPPA-compliant measurement, Singular and Kidoz can provide campaigns that are tabulated and attributed properly for clients & publishers.

Kidoz Co-CEO Eldad Ben Tora recently sat down with Singular’s John Koetsier, to talk about kid-safe marketing and the potential for app install campaigns on the Kidoz network with Singular attribution enabled.  The interview in its entirety can be viewed here.

 

“The demand for kid-safe attribution is rapidly growing,” said Eldad Ben Tora, Co-CEO of Kidoz. “Previously app owners that were looking to optimize their user acquisition budget were limited in their ability to track the value of different traffic sources.  Via our partnership with Singular, every advertiser can now correctly attribute installs and understand the value of each source without compromising the privacy and security that users, advertisers, and publishers expect from Kidoz.  Singular is a leader in marketing analytics and attribution and this partnership expands the possibilities of kid-safe marketing.”

“Singular is committed to enabling the best marketing analytics for our customers and connecting to the leading COPPA network Kidoz with our COPPA-compliant measurement solution is a perfect match,” said Susan Kuo, Co-founder & COO of Singular.  “Connecting with Kidoz provides Singular customers with a new option for reaching children and families in a secure and private way and we look forward to building our business relationship together.”

About Singular

Singular (www.singular.net) is a marketing analytics and attribution platform that unifies marketing data, giving marketers actionable insights from previously siloed data sets.  By connecting upper funnel marketing data with lower-funnel attribution data, marketers can measure ROI from every touchpoint across multiple channels and optimize spend down to the most granular levels.  Singular currently tracks over $10 billion in digital marketing spend to revenue and lifetime value across industries including commerce, travel, gaming, entertainment, media, and on-demand services.  Singular is backed by Norwest Venture Partners, General Catalyst, Thomvest Ventures, Method Capital, Translink Capital, DCM and Telstra Ventures.

About KIDOZ INC.

Kidoz Inc. (TSXV:KIDZ) (www.kidoz.net) owns the leading COPPA & GDPR compliant contextual mobile advertising network that safely reaches hundreds of million kids, teens, and families every month.  Google certified and Apple approved, Kidoz provides an essential suite of advertising technology that unites brands, content publishers and families.  Trusted by Disney, Hasbro, Lego and more, the Kidoz Contextual Ad Network helps the world’s largest brands to safely reach and engage kids across thousands of mobile apps, websites and video channels.  The Kidoz network does not use location or PII data tracking commonly used in digital advertising.  Instead, Kidoz has developed advanced contextual targeting tools to enable brands to reach their ideal customers with complete brand safety.  A focused AdTech solution provider, the Kidoz SDK and Kidoz Programmatic network have become essential products in the digital advertising ecosystem.

The Private Securities Litigation Reform Act of 1995 provides a “safe harbor” for forward-looking statements.  Certain information included in this press release (as well as information included in oral statements or other written statements made or to be made by the company) contains statements that are forward-looking, such as statements relating to anticipated future success of the company.  Such forward-looking information involves important risks and uncertainties that could significantly affect anticipated results in the future and, accordingly, such results may differ materially from those expressed in any forward-looking statements made by or on behalf of the company.  For a description of additional risks and uncertainties, please refer to the company’s filings with the Securities and Exchange Commission.  Specifically, readers should read the Company’s Annual Report on Form 10-K, filed with the SEC on March 31, 2021, and the prospectus filed under Rule 424(b) of the Securities Act on March 9, 2005 and the SB2 filed July 17, 2007, and the TSX Venture Exchange Listing Application for Common Shares filed on June 29, 2015 on SEDAR, for a more thorough discussion of the Company’s financial position and results of operations, together with a detailed discussion of the risk factors involved in an investment in Kidoz Inc.

For more information contact:
Henry Bromley
CFO
ir@kidoz.net
(888) 374-2163

In the past few years it seems that every Apple event caused massive ripples across the advertising landscape. This year’s WWDC 2021 was no exception!

As you know by now, the most interesting news for us is that in iOS 15, app developers will be able to define an endpoint that will receive a copy of all the winning SKAdNetwork postbacks (there’s also iCloud+ Private Relay, but that will be a separate post).

This change is very welcome, and we’re excited about it since it creates better transparency for advertisers, but if you think about it more carefully – you realize it can also add complexity, and generate some interesting implications for ad networks and advertisers.

Like every Apple change, this is another part in the ongoing shift in our reality. And I actually think I found the perfect GIF to describe the last year:

The good: Direct access to SKAdNetwork postbacks in iOS 15

More visibility is always a good thing. And getting data right from the source is dramatically better than what we had before. It’s also great to see Apple starting to acknowledge the advertisers in the equation (it felt like SKAN was built for ad networks more than it was for the people who pay the bills)!

Some of the notable benefits of this change that Singular customers will get:

1. Access to ALL the winning postbacks, including those that were sent to Facebook and Google, in our API, raw logs, or ETL straight to your database.
2. Tools to analyze your privacy thresholds, and how often data is redacted.
3. Full visibility into publisher level reporting. Do you remember the days where networks were masking the publisher who drove the install? Now Apple is canceling that, and making the identity of the publisher app transparent (as long as privacy thresholds are met, of course).
4. Ability to pass information on top of the conversion value that your ad networks don’t necessarily need or understand. For example: 3 bits could be used for event mapping (tutorial complete, social login, purchase), and 3 bits could be used for day of install.
   The ad network can use the event bits, but Singular would use the full 6 bits.
5. Ability to compare network vs SKAN numbers in Singular: We ingest aggregate reporting from networks and decoded SKAN postbacks from networks and raw SKAN postbacks from your app. We can now compare all three and verify the numbers are accurate … and more importantly, that the conversion values are accurate.
6. Multiple layers of postback decoding:
   1. Level 1: Raw postback as received from the app
   2. Level 2: Decoded conversion values
   3. Level 3: Decoded skan_campaign_id information
   4. Level 4: Matching of a SKAN postback to upper-funnel campaign/creative data, with spend and ROI
7. Better protection against fraud: With iOS 14.6 the SKAN postbacks are still sent from the device, but they go through a proxy Apple is running. With iOS 15, there will be 2 copies sent of the winning postback: one to Singular, and one to the network. This helps us detect any manipulation that happened outside of the device. (I have to say that a motivated fraudster can still exploit SKAN since it’s not a perfectly secure setup, and I don’t want to share all the secret sauce, but we’re building proprietary mechanisms to detect such exploits.)

Wishlist: maybe Apple is going to start sending organic install postbacks as well with conversion values so that organic reporting via SKAN could work as well? Pretty please!

The bad: these postbacks, on their own, are limited in their value

As I’ve described above, getting postbacks in iOS 15 is unequivocally a good thing.

BUT, these postbacks are no panacea… in fact, far from it.

Due to how SKAdNetwork works, the advertiser has no control over the SKAN Campaign ID (reminder: 100 possible values). These are controlled by the ad network, as they should be, and each ad network’s mapping is vastly different from the next.

Since advertisers do not and probably will not have access to this mapping, it means that these postbacks still don’t provide a functional way to build reporting off of them alone. We still need to go to all the other scattered sources to assemble functional reporting.

To illustrate, the data coming from these postbacks will look like this:

Ad network	Source App	Campaign ID	Conversion Value
Facebook	Facebook	7 -> ???	21 -> (Revenue $10)
Google	Google	53 -> ???	36 -> (Revenue $17)

It will be interesting to see if ad networks will provide the mapping to the SKAN campaign IDs. My guess is that they won’t share it, for the following reasons:

1. Facebook and Google probably use some sophisticated machine learning model to associate their campaigns, ad groups/ad sets, and creatives to the “SKAN Campaign ID,” and it won’t be trivial to share it.
2. Since SKAN postbacks have random timers, you can’t always tell what mapping was used for a given SKAN postback. That means that some of these mappings might be statistical by nature, which is not something that is easily sharable with advertisers.
3. If Facebook and Google won’t offer it, I think other networks won’t offer it either. It’s too much work, and they already offer translated postbacks through the pipes they built.

The ugly: The crazy complexity marketers face in iOS app marketing

When you have one good clock, you know exactly what time it is. When you have two that vary, you’re not sure anymore. Three? Good luck.

A single — and accurate — source of marketing truth functions the same way. But now mobile marketers have no fewer than six, and probably more.

1. IDFV data for installs (but of course without attribution information)
2. IDFA data for a fraction of installs (with full attribution information)
3. Raw SKAN postbacks from Apple (in iOS 15, and for many ad networks that work with Singular, already now in iOS 14)
4. Translated postbacks from networks decoded to what they mean (to the networks that share that)
5. Aggregated reporting from Facebook, Google, and other networks that include SKAN data, IDFA opt-in data and their modeled data (the only place with actual campaign names, creative names, impressions, clicks, cost, etc.)
6. Deeplink and owned media data

How many clocks do you have now? Which ones are accurate? Do you need to pay attention to all of them? Now, clearly, the one clock that you trusted may not have been right. But your life certainly gets more complicated with six or seven clocks.

So now a marketer wonders: what do I do? Do I pick a few of them and optimize based on those? Do I turn to media mix modeling? Do I turn to incrementality testing? Are these signals enough?

And of course, some marketers start wondering again about fingerprinting. Seeing some MMPs offer it as the default option (!!) despite being non-compliant, or mask it as something else, makes people confused about whether they can just ignore all of this SKAN complexity and … fingerprint. I recently talked to a major mobile app developer who’s done a phenomenal job getting ready to adopt SKAN, and actively adopting it, and he felt he’s the only boy scout left, since some competitors are fingerprinting their installs.

Plus, of course, there are still significant gaps in SKAdNetwork.

Conversion values should be signed, so there is a reduced risk of fraud (although even that would not close that loophole completely). We desperately need a SKAN solution for web-to-app.

The reality: marketing measurement is changing

We are reinventing the future of marketing measurement in real-time: fixing the plane as we fly it.

We need SKAN to be sophisticated. Way more than it is now. We also need to incorporate other complementary solutions (like media mix modeling and incrementality), but the important realization is that there are going to be multiple signals coming into play, and not a single source of truth like we had with IDFA.

That’s why whenever someone asks me about the IDFA changes, and impact to our customers and our business, I tell them that the need hasn’t changed, but the technology did, and MMPs like Singular have an amazing opportunity to build the next-gen measurement that incorporates all available signals.

And we need to produce a simplified hybrid view of marketing effectiveness reality that takes in all the available information on inputs and outputs to give growth teams the directional insight they need to optimize campaigns.

I’ve talked to multiple marketers in the past weeks and months who almost feel exhausted. They’ve made so many changes over the past year and they see more on the horizon.

Our job at Singular is to do that work for you.

And, as we always have, provide the insight you need for growth.

Feeling like Inception is happening to you?

We’re happy to chat. Join our Mobile Attribution Privacy channel on Slack, where there are thousands of smart marketers who can help, and where you can message me directly. Or ping us for a quick meeting to assess how we can help.