Recent reporting from the FT highlights Apple’s growing ambitions with its advertising business, noting the large number of open positions that Apple is currently promoting for its advertising platform. I argue in Apple robbed the mob’s bank that Apple’s ability to expand its advertising business is a direct consequence of its App Tracking Transparency (ATT) privacy policy, which diminished the fortunes of hub-and-spoke advertising platforms like Meta and Snap and created the opportunity for Apple’s platform to gain share.
In the second installment of the series, I cite analysis from the research group at Evercore ISI which estimated that Apple’s advertising business could grow to as large as $25BN in annual revenue by 2025 as a result of ATT. Per the FT, that number has been updated, and Evercore now projects Apple’s annual advertising revenue to grow to just over $30BN by 2026.
An interesting thought experiment to conduct in the context of these numbers is: how could Apple expand its current advertising business to reach that scale by 2026? Or, put another way: what inventory does Apple sell over the course of 2026 for its annual advertising business to account for $30BN in revenue?
Currently, Apple’s Apple Search Ads platform, which was launched in 2016, only services app install ads across inventory within the App Store app (notably, with new placements), Apple’s Stocks app, and Apple’s News app. Apple revealed at the beginning of this year that the App Store sees 600MM weekly visitors, and in April 2020, Tim Cook announced in an earnings call that the Apple News app had reached 125MM monthly active users.
Mobile market intelligence company Data.ai appraises the total mobile advertising market, across both app install ads and mobile web ads, at $295BN in 2021. And Appsflyer, the mobile advertising measurement company, estimates that app install advertising accounted for between $78 and $83BN in 2021, or between 26% and 28% of total mobile advertising spend for the year (per Data.ai).
Further, SensorTower, another market intelligence company focused on mobile, estimates that global consumer spend in apps will grow by a 12% CAGR between 2021 and 2026 to a total of $233BN, with App Store (iOS) revenues accounting for 69% of the total in 2026.
It’s possible to project total app install spend to 2026 using a very rough set of assumptions, garnered from different sources that utilize different methodologies:
- Total mobile advertising spend will grow at 12% CAGR from 2021-2026, consistent with SensorTower’s estimate of consumer in-app spend growth over that period (in other words: advertising spend growth is equivalent to consumer spend growth);
- The share of app install advertising spend relative to total mobile advertising spend will remain constant at 27% from 2021-2026, using the midpoint of Appsflyer’s estimate for app install advertising spend in 2021 ($80.5BN) and Data.ai’s estimate for total mobile advertising spend in 2021 ($295BN). Total mobile advertising spend is backed out of app install advertising spend, which increases by 12% per year;
- The share of app install advertising spend for iOS is consistent with SensorTower’s estimate of consumer in-app spending by platform per year.
Using these (admittedly, tenuous at best) assumptions, total mobile advertising spend can be projected to reach $520BN in 2026, with $142BN of that being allocated to app install advertising, of which, $98BN is deployed to iOS app installs.
In order for Apple to achieve $30BN in advertising revenue in 2026 from app install advertising alone, it would need to capture roughly 31% of the overall iOS app advertising market. Using the aforementioned estimates of total app install advertising market size ($90.2BN, of which $62.3BN is deployed on iOS) and Apple’s 2022 advertising revenue ($5BN), Apple will take roughly 8% of the iOS app install market in 2022.
Can Apple nearly quadruple its share of the app install market by 2026?
It’s unclear if that can be accomplished through owned inventory alone, given that the totality of the combined impact of ATT and new ad placements in the App Store has only resulted in an 8% share of market. It’s possible that additional placements in the App Store, such as the new hero placement in the Today tab, increase ad load by a greater degree than net new reach, since ads already exist there. And while the Maps and Podcasts apps are popular, it seems safe to assert that both will be less effective at delivering app install ads than the App Store.
But Apple is rumored to be building a DSP that might traffic inventory in third-party apps: doing so would allow Apple to aggregate inventory from across the iOS ecosystem and to serve those placements through its own ads targeting infrastructure, which utilizes behavioral data — app installs and in-app purchases — to score ad relevancy, as I outline here. Expanding its scope outside of its owned and operated inventory could dramatically enlarge the surface area of Apple’s ads business. (Note: I use the term platform to reflect an advertising system that only operates on owned inventory, and network to reflect a system that aggregates and serves third-party inventory)
Recent speculation around Apple’s ad network ambitions stems from job postings related to such a system, but clues that Apple intends to build an ad network have existed for a number of years. Back in May of 2020, before Apple had announced its ATT privacy policy, I explored Apple’s potential motivations for building an ad network that services third-party inventory in a post aptly titled, Why is Apple building an ad network?. The post was a reaction to Apple’s addition of two new parameters to its ad platform’s campaign management API: supplySource and adChannelType. These new parameters implied that Apple might introduce third-party sites to its ad platform, setting to stage for Apple to expand its advertising business to traffic third-party inventory. From the piece:
Note that Apple also potentially has a competitive advantage in serving ads within its proprietary apps and more broadly throughout its iOS ecosystem: it has direct access to device and advertising identifiers that may soon be deprecated and no longer accessible to advertisers. If Apple does deprecate the IDFA advertising identifier, what ad network would be better positioned than ASA?…If Apple grows ASA into a bonafide ad network, and especially if Apple does deprecate its advertising identifier, Apple could propel ASA into the top echelon of mobile app install advertising platforms.
But Apple is not constrained to its current operational domain of app install ads. Apple will almost certainly integrate advertising placements into more of its free apps — such as Maps, Podcasts, and Books — and could potentially create ad-supported tiers for subscription services like Apple TV+, Apple News+, and Apple Arcade. These placements might facilitate goals other than direct-response app installs: they could support brand awareness (Apple TV+) or brick-and-mortar retailer visits (Maps) or delayed-response conversions (Podcasts), or some combination of all of these in addition to app installs.
The true prize of an exanded advertising business would be a retail and ecommerce offering. ATT impaired mobile ecommerce advertising — which was Meta’s largest advertising vertical — but Apple couldn’t absorb any of the spend that fled other channels in search of yield because its advertising platform only delivers app install ads. If Apple was able to accommodate ecommerce advertising spend, it’d be the natural home for budget that was cut as a result of ATT. Apple’s targeting data is currently sourced from iTunes, and it therefore only relates to app engagement. But there’s no reason Apple couldn’t use its other first-party pools of user-level data for targeting ads to web-based purchases instead of app installs (although Apple does say explicitly in its privacy documentation that Apple Pay data is not used by the ad platform).
Edit: a previous version of this article referred to a report from a mobile marketing analytics company. I have removed the reference to the report and adjusted the relevant analysis.