The impact of Apple’s IDFA update for our ecosystem

The impact of Apple’s IDFA update for our ecosystem

This Monday, Apple shared news on the way how the IDFA is handled in iOS 14. Similar to many other privacy-related data, there will be a specific notification popping up to ask the user for consent to use the IDFA. This is happening now for each app compared to a global setting hidden in the privacy settings (Limit Ad Tracking - LAT). The text for the notification is customizable, but there won't be the same flexibility in consent design like we see on the Web for GDPR (i.e. no UX tricks possible to increase Opt-in ratio).

No alt text provided for this image


If the user is not giving any consent, the IDFA will be zeroed (same as an enabled LAT setting), so it can't be used for any user-based cross-app mechanics (i.e. deterministic attribution, lookalike targeting, profiling, retargeting, user segmentation). Almost the complete app ecosystem uses the IDFA to deliver value to marketers - an IDFA based advertising mechanic almost always outperforms a non-IDFA approach. The reasons are obvious:

  1. IDFA right now is a persistent valuable data source and most machine-learning approaches use this as the main input
  2. Measurement and deterministic attribution is possible, a marketer knows what performs where

It didn't make sense to build advertising tools not being based on IDFA as such products would've been inferior. Even user acquisition tools use the IDFA as input for optimization to learn which users are installing an app (and spending money). With no IDFA as an input the whole ecosystem has to rethink their approach:

  1. User-based targeting and optimization will not be possible. The context will become the main input (i.e. publisher, creative)
  2. Deterministic attribution will not be possible. Apple shared in a workshop yesterday that they try to kill all alternative fingerprinting approaches, too (IPv6, MAC). Attribution will be owned by Apple via their SKAdnetwork, but it's not user-based and not real-time and they will only be reported if the group of converters for a campaign is big enough.

Actually, we have already around 30% of iOS inventory right now, where this is not possible: The inventory where users have switched on LAT. How is this going to continue now for our industry? How will the market shift for Ad Networks, DSPs, Attribution Providers, etc? Let's dive into some numbers:

These are all rough estimates, advertising share for iOS is very likely higher than the 25% OS penetration (more like 35-40%), so let's adjust accordingly. The industry impact on how the inventory will shift to non-ID is as follows:

October 2020: 14% of total market impacted (35% iOS ad market share * 50% iOS 14 penetration * 80% opt-out)

March 2021: 25% of total market impacted (35% iOS ad market share * 90% iOS 14 penetration * 80% opt-out)

While this doesn't look much on the surface, we are talking about a $60 Billion market impact (based on App Annie's State of Mobile 2020 report)!

If you would turn it now into a positive light, we are now having a $60 Billion market opportunity with completely new products to emerge.

The users will still be there, the advertisers will be there.

The ad impressions will be there, the billions of ad spend will be there.

Now let's reinvent the future of our industry.

Felix Gillen

Wegbereiter für Innovationen 👨🚀👋

4y

I guess there will always be some way to track using fingerprints for mobile attribution even if apple tries to prevent this.

Raphael Haase

Stay hungry, stay foolish | Innovator for greenfield projects in medical and finance | All-rounder from business requirements over architecture to implementation

4y

If Apple is successful and more companies follow suit, this will be a huge change for most ad companies and a disaster for some. But I think it is in the interest of the consumer. There should be less distractions and less utilization of private data. And there should be taxation of consumption combined with the complete abolishment of taxation on income/profits. We all must consume less and be less distracted.

Rocco Strauss

Internet Equity Research Analyst; Partner at Arete Research

4y

Thanks for sharing your thoughts @Pan K.. I have a few questions/comments I would love to get your view on: (1) Could app developers push alternative identifiers through the RTB pipes (email addresses, phone numbers, etc.), i.e. users are going to see a similar push as by the web publishing world, that asks for logins after now 3-5 free articles, or similar. (2) Would Facebook Single Sign On (SSO) be a good alternative, thus FAN should actually see a push and not be wound down? (3) App developers could prompt the opt-in notification every time a user opens the app and most "average users" don't care about privacy and will want to get it away. (4) The last draft of the ePrivacy regulation makes clear, that app developers can make the opt-in a requirement for the app to be used, so apps with utility (whatever that means for each user individually) are likely having an easier way to obtain consent. (5) Could Apple aim to develop a new "iAd" version? Guess there "app of the day" ad format, which is highly targeted within the app store would also suffer. (6) App developers could be less incentivised to build apps for Apple/Android. Another chance for Facebook to push VR or its super-app structure? (7) $60bn up for grabs ... a lot of that spent is highly targeted, so afraid the larger walled gardens could be the beneficiaries ... similar to GDPR.

Carsten Frien

Founder & CEO at Roqad - I am hiring!

4y
Like
Reply
Carsten Frien

Founder & CEO at Roqad - I am hiring!

4y

Pan, great article, as always! Especially the $60Bn "opportunity view" you take. We at #roqad have been anticipating this move for some time and have brainstormed how to tackle this problem (and the 3P Cookie problem). Anyone interest to learn how our identity graph product can help today and in the future please leave a note here or ping me via DM

To view or add a comment, sign in

Insights from the community

Others also viewed

Explore topics