Are you prepared to survive the IDFA Apocalypse? Use this checklist to grade your progress.
To customize your list, which solution
do you use for ad measurement?
Last updated: January 7, 2021
At WWDC 2020, Apple threw the mobile industry into chaos by announcing plans to deprecate the IDFA (Identifier For Advertisers) on iOS. These changes are expected to arrive in early 2021.
Our mission at Branch is to help mobile teams everywhere grow their apps. We've compiled this checklist to help you understand whether you're on track, and uncover any gaps in your plans.
Set up a time today to learn more about preparing your marketing stack.Request a Demo
Here's a list of all the steps you've already taken, and more info on areas you might still want to check!
The best place to start is by making sure you're in the loop on the latest news. Which of the following have you already completed?
There is plenty of spin in the ecosystem around these new iOS 14 privacy policies, with vendors and independent thought leaders publishing various interpretations of what they believe Apple is trying to do.
Interpretations are all well and good (and often quite useful!), but it's also worth reviewing the details straight from the source.Read More
SKAdNetwork is Apple's attribution framework, facilitating ad measurement without device-level data. It is a fundamental change to how ad attribution works on iOS, and some of the biggest changes include:
If you have not already read up on SKAdNetwork, now is the time! You can also reach out to us at Branch and we'll be happy to give you a full run-down.Read More
We have built a suite of new functionality to make sure you're set up to get the most from SKAdNetwork, including dashboard reports, SDK updates, and updated integrations with ad networks.
As of early January 2021, the necessary iOS SDK updates are already available for all customers, and we're in the process of rolling out the new dashboard configuration and reporting pages to Universal Ads subscribers.Read More
Take a few minutes to understand the broader impact of iOS 14, and what changes (if any) you can expect to non-attribution areas of your Branch implementation.Read More
Savvy marketers are going to find that SKAdNetwork data alone is insufficient for their needs. Branch's Predictive Modeling engine is designed to operate in parallel to SKAdNetwork, using privacy-safe event analysis to provide granular insights that allow you to continue optimizing your campaigns.
For savvy marketers, SKAdNetwork is not a satisfactory replacement for the granular campaign insights that IDFA-based measurement provided. Some MMPs are planning to offer some sort of 'probabilistic' fallback data, but these methods vary widely in both coverage and accuracy.
Here are the two main questions you want to ask:
For savvy marketers, SKAdNetwork is not a satisfactory replacement for the granular campaign insights that IDFA-based measurement provided. Mobile Measurement Partners (MMPs) — including Branch — offer solutions designed to work in parallel to SKAdNetwork, using privacy-safe event analysis to provide granular insights that allow you to continue optimizing your campaigns.
To learn more, reach out to your Branch contact, watch our Navigating IDFA Changes webinar, or set up a time to talk. If you're comparing solutions, keep in mind that these methods vary widely in both coverage and accuracy. Here are the two main questions you want to ask:
Before you jump into implementation steps, there are a few important choices to make first.
Adapting successfully will require support from multiple teams in your organization. Here are a few key stakeholders to include:
Unless you have an extremely simple (and short!) customer lifecycle, a basic SKAdNetwork implementation is not going to provide all the data you need to measure the long-term performance of your acquisition campaigns.
There are a variety of ways you can approach this, ranging from straight-forward (using a system like Branch's Predictive Modeling in parallel with SKAdNetwork, essentially a more privacy-safe continuation of the pre-iOS 14 status quo), to adventurous (building an econometric analysis model to show campaign incrementality).
While collecting IDFAs might seem like one simple way to retain device-level attribution data, Branch believes the IDFA is functionally dead for most use cases (users aren't typically going to select 'Allow App to Track' for a new app they just downloaded).
However, requesting it may still make sense for certain apps/verticals, especially if you can show a clear value proposition to users (though take care to understand Apple's guidelines on what constitutes 'unacceptable' incentivization).
If you decide to collect IDFAs, the
AppTrackingTransparency modal is a permission prompt. And as with any permission prompt (just think of push notifications or location info for a few examples), there is both an art and a science to securing the user's consent. You'll need to consider things like pre-permission prompts and testing when in the onboarding experience to trigger the pop-up.
One thing you shouldn't do right now: implement the official
AppTrackingTransparency modal before Apple requires it. Because Apple isn't yet enforcing the new permission, you could end up unintentionally 'opting in to the opt-in' earlier than necessary…and losing critical insight right as it's most important for understanding all your other preparations. Instead, if you plan to request the permission, you can begin by experimenting with a 'soft prompt' that you control yourself.
conversion-value parameter allows you to get a limited amount of post-install insight, but it has significant technical limitations (by design).
MMPs (including Branch) have built functionality that uses existing SDK integrations and dashboard-based configuration tools to simplify the management of
conversion-value codes. This will likely be sufficient for many advertisers, at least as a starting point while the ecosystem learns how to work effectively with SKAdNetwork. However, if you want to build a more custom model, you can opt out of your MMP's functionality and implement the
conversion-value functions yourself.
Either way, it's important to invest some time now to consider your current customer journey and pick which subset of events will provide meaningful performance insights within the limits of SKAdNetwork.
These are the important in-app and infrastructure updates you'll need to make.
If you decide to collect IDFAs, you'll need to implement the
AppTrackingTransparency framework. This is Apple's new system for requesting user permission to collect data (including the IDFA) for the purpose of cross-app tracking.
You'll most likely need to integrate this yourself, because third-party vendors (including Branch) are generally not planning to trigger this permission on your behalf via their SDKs (though Branch will automatically collect the IDFA for you if you've received permission from the user).Read More
Branch supports two options for SKAdNetwork integration:
For most advertisers, the first option is simpler and preferred.
Note: if you are using a server-to-server integration, you must integrate the calls directly. Reach out to your Branch contact for more info.Read More
Every MMP has released updated SDKs with support for in-app implementation of the SKAdNetwork framework. Review the integration steps for your MMP:
You can also disable the built-in support and integrate the
updateConversionValue() SKAdNetwork calls into your app directly.
Note: if you are using a server-to-server MMP integration, you must integrate the calls directly.
If you are not working with a MMP, you can integrate the
updateConversionValue() SKAdNetwork calls into your app directly.
conversion-valuedefinitions without continual code changes and App Store updates
You'll probably want to tweak your SKAdNetwork settings regularly, and you won't want your marketing campaigns to be blocked by your product release cycle. Make sure your team has a way to update
conversion-value definitions without re-architecting your code-level event tracking implementation each time.
iOS devices send SKAdNetwork conversion postbacks directly to each registered ad network. Unless you run ads via only a small number of partners, it will quickly become unwieldy to check each network's reports separately.
You'll want an automated workflow to collect this data and aggregate it for analysis (pro tip: this is where a mobile measurement partner like Branch typically fits in to your stack).
Most in-house systems should already be using other identifiers (for example: IDFVs), but it's a good idea to do an audit to make sure nothing internal will break when IDFAs stop coming through.
+match_guaranteed parameter provides a way to securely implement data-sensitive use cases (for example: auto-login, PNR visualization).
+match_guaranteed will usually be false on install events (there will be no change for non-install events), unless users opt-in to IDFA collection.
setIdentity()with your Branch SDK integrations
Sending unique, anonymous identifiers for your users was a great best practice even before IDFA deprecation. Without IDFAs, setting these identifiers (and doing so as early in the user journey as possible) will make it easier for you to analyze your data, and will help Branch measure cross-platform, cross-device user journeys.Read More
While this checklist is primarily focused on preparing for iOS 14 as an advertiser, if you also monetize your app using ads, you'll likely also need to update those integrations to work with SKAdNetwork.
Check with your ad network partners for more information.
Aside from technical updates, there are also a few other things to take care of before you're ready to go.
We're in the process of rolling out our new SKAdNetwork configuration and reporting dashboard functionality to all Universal Ads customers. In the meantime, you can reach out to your Branch contact or email support to request early access.
If you're planning to use the Branch SDK's built-in handling of
updateConversionValue() SKAdNetwork calls, you need to enable this via your Branch dashboard.
This is important, because you only want one party firing these calls. If you integrate them directly (or use another third-party to do so), you don't want accidental conflicts and should leave this disabled.Read More
conversion-valuemappings in the Branch SKAdNetwork dashboard
If you're using Branch's built-in handling of SKAdNetwork calls, this process will set up the mapping between existing Branch tracked events and when the Branch SDK fires SKAdNetwork
If you're making the SKAdNetwork calls yourself, this will still allow you to associate human-readable names to the raw
conversion-value integers, for easier reporting.
If you're planning to use your MMP's built-in handling of
updateConversionValue() SKAdNetwork calls, you need to enable this via your MMP's dashboard.
conversion-valuemappings in your MMP's SKAdNetwork dashboard
If you're using your MMP's built-in handling of SKAdNetwork calls, this process defines when the MMP's SDK will fire SKAdNetwork
This is important, because you only want one party firing these calls. If you integrate them directly (or use another third-party to do so), you don't want your MMP's SDK getting involved and causing accidental conflicts.
Most ad networks are planning to use SKAdNetwork on iOS, but many of them are also supporting device-level attribution data in parallel (through Branch's Predictive Modeling system, or basic probabilistic methods from traditional MMPs).
Branch is maintaining a list of networks and campaign types that support each method, but it is a good idea to confirm with your networks directly.
Facebook plans to use SKAdNetwork for ad campaigns targeting iOS 14 users, and has released additional information about how iOS 14 will affect advertisers. This includes a
conversion-value mapping process, to allow Facebook to make campaign optimizations based on post-install user activity.
If you're working with an MMP, the attribution and measurement components (including the
conversion-value mapping) should be handled behind the scenes on your behalf, but you will still want to be aware of the limitations on ad accounts and total campaigns per app.
Facebook is also making changes to web ad campaigns on iOS 14, including deprecation of longer attribution windows and limits on the total number of conversion events per domain.
If you run web campaigns on Facebook (or use app attribution links inside web ad units), you'll want to get familiar with these changes too.Read More
AdServicesframework for Apple Search Ads
Apple Search Ads campaigns will not use SKAdNetwork for attribution, but Apple is still planning to make changes to align with
If you use an MMP to manage Apple Search Ads attribution, these updates should be handled behind the scenes on your behalf when they eventually roll out.Read More
Other major networks (Google, Twitter, Snap) have not yet publicly communicated advertiser updates for iOS 14. We expect to learn more details in early 2021.
Apple has already enabled the new privacy report section of App Store product pages. You'll need to complete the Privacy Practices Questionnaire before submitting your next app update.
If you're using Branch, you can find the information you need about which data types we collect on your behalf here.Read More
Once you've completed all the critical preparation items, here are a few longer-term projects to consider.
The iOS 14 privacy changes primarily affect paid ads, but ads aren't the only mobile marketing option. This is a great time to consider new ways to leverage both owned and earned channels, including user content sharing, organic social media, web smart banners, and deep linked email campaigns.Read More
Get creative! Build attribution in web campaigns and use deep linking to drive users straight to the right content inside your app. If you instead need to drive users directly to your website because of the way the ad partner operates, use a Branch link configured to route users via your website, and still get attribution once the user eventually lands in your app.Read More
Most savvy marketers are going to find that SKAdNetwork data alone is insufficient for their needs. Branch's Predictive Modeling engine is designed to operate in parallel to SKAdNetwork, using privacy-safe event analysis to provide granular insights that will allow you to continue optimizing your campaigns. Traditional MMPs typically offer some sort of 'probabilistic' fallback that sounds similar on the surface, but these methods vary widely in both coverage and accuracy.