Matt Levine, in his excellent newsletter focused on all matters finance, Money Stuff, coined a motto: “Everything is Securities Fraud.” Levine expounds on the catchphrase in this podcast episode with Preet Bharara, but in short: because public companies are ultimately beholden to their shareholders, any malfeasance by a public company’s management or board can be considered a form of securities fraud.
If I was forced to choose a motto for Mobile Dev Memo, at least for this very specific moment in time in which the digital advertising landscape convulses in a state of momentous change, it would be: “Everything is an ad network.”
Not literally everything. But any company with sufficient supply of or access to consumer data — in a first-party environment! — is now presented with the opportunity to build an advertising network where previously that would have been impractical or even somewhat absurd given the competitive landscape. As I wrote in 2017 with Facebook and Google own our eyeballs:
That ownership of engagement leads to dominance in advertising isn’t very remarkable — Facebook and Google aren’t going to let 3rd parties sell their native inventory, and with Audience Network Facebook has actually managed to do just the opposite. But what will be interesting to track over the next 12-18 months is how this dynamic changes as Facebook and Google turn their individual apps into content and app distribution portals. Facebook’s efforts here have been fairly obvious and well documented, but Google’s have been furtive: ordering an Uber directly from Google Maps, for example. Google’s AMP project (which isn’t utilized within an app but is nonetheless helping to drive mobile advertising revenue growth for the company) and Facebook’s Instant Articles product are also shifting ad revenues away from other players.
As I point out in that article, Facebook and Google collectively captured 89% of all digital advertising revenue growth in the US in 2016. These companies thrived at a time when bigness and momentum bestowed advantage on the hub-and-spoke model of digital advertising: operate what is essentially a data warehouse of behavioral signals, convert that data into targeting parameters, apply those targeting parameters to owned-and-operated ad inventory, and then create a positive feedback loop for advertisers by ingesting all of their relevant engagement data in what I’ve called the events stream. I detail this process here, and while it worked, it worked beautifully.
But it doesn’t work anymore. The policy-based change agents I cite above are undermining digital advertising’s hub-and-spoke model: these changes bestow a privilege upon any form of first-party data that can be used for ad targeting in a native environment.
Facebook’s first-party data — its knowledge of your likes and comments, the groups to which you belong, etc. — is not very helpful for ad targeting. Or it’s at least less helpful than its knowledge of the fact that you purchased gems in a mobile game or a pair of sneakers on an eCommerce website. But Facebook’s ability to observe those types of purchases is being diminished. Many of the companies that previously transmitted their purchase data to Facebook and other ad platforms very enthusiastically are determining that their data can now be used to power a proprietary ad network.
The list of such cases is extensive:
- Lowe’s, the home improvement store, has launched an ads business (Home Depot launched one last year);
- Retailer Target revamped its ad network in 2019, and Walmart has recently expanded its ads business through partnerships and M&A;
- Instacart started an ad network last year and is investing heavily into that effort, including through appointing Fidji Simo, the architect of Facebook’s mobile ads platform, as its CEO;
- Doordash launched an ad platform this year, following UberEats’ decision to build an ad network in 2019;
- Zoom, the video conferenceing software, is introducing ads to its free tier of users, although it’s not clear that this will be powered by a proprietary ad network;
- Examples from gaming are too numerous to fully list, but: Zynga, the mobile games developer, acquired Chartboost, a mobile ad network; Skillz, an in-game wagering platforrm, acquired Aarki, a mobile DSP; hybrid game developer and ad platform Applovin acquired MoPub; ironSource, which similarly operates a portfolio of mobile games and an advertising platform, acquired ad network Tapjoy;
- Pharmacy operator CVS launched an ad network in 2020, as did Walgreens.
Each of these companies sees an opportunity to layer ad impressions atop first-party data so as to facilitate proprietary ad targeting. This opportunity is new: while Google and Facebook took 89% of all ad spend growth previously because their enormous reach provided them with the leverage and ability to ingest third-party data from advertising clients, now those advertisers are wrapping their arms around their data and monetizing it. Apple may be the primary beneficiary of this policy change: its own ad network’s revenue has been projected to reach $20BN by 2025 from $2BN in 2021 as a direct result of ATT.
It’s understandable that any company with sufficient size and scale of data would want to build an ad network. Ad networks are highly profitable, generally running on very thick gross margins and mostly monetizing content and product surface area that already exists. So as the norms for data collection and aggregation change, favoring first-party contexts, the companies that can package together their first-party customer data to attract advertising revenue will do just that. This aligns with my Content Fortress thesis: the primacy of first-party data will drive consolidation and horizontal integration such that any pool of content can be yoked to advertising infrastructure to lucrative effect.