AppLovin is driving people mad

Everyone’s favorite advertising company, and stock, drives people crazy. Is it the most innovative company of our era, or a total fraud? Or both?

The AppLovin debate

Another week, another short seller report on AppLovin. Despite its name, $APP ( ▲ 2.94% ) is a company people love to hate. Let’s step back and assess.

Why do we care?

The rise of AppLovin is fascinating because it embodies many “narrative violations” that are difficult for industry observers to square.

Some of those:

• The mobile app market is dominated by Meta, Google and Apple. Independents cannot survive.

• You can’t build a performance business in “open” environments, you need proprietary data or O&O inventory.

• There aren’t any success stories in mobile app, all the DSPs and SSPs died.

• Unless you’re The Trade Desk $TTD ( ▲ 0.8% ) you can’t succeed as a public ad tech company.

Oh well, I guess the conventional wisdom was wrong. That makes a good story.

APP’s distinctive M&A history

I first did business with AppLovin during my time at Beeswax, when they were in a peer group with IronSource $IS ( ▼ 2.83% ) and Unity $U ( ▼ 1.47% ) , companies it has since eclipsed. (As a side note, our engineering team thought the company name was so funny that I ended up registering the domain bidlovin.com, which I still own…make me an offer)

AppLovin did three fairly brilliant things to break away from the pack. First they acquired Jim Payne’s MAX, in an unheralded deal which put the company in the pole position for the transition from mediation to unified auctions (i.e. waterfall to header bidding for those of you from the web world).

Next, they built a gaming studio called Lion Studios and invested in various other gaming companies. This gave them the O&O sandbox where they could a) collect enormous amounts of data; b) train their algos on what really works for app acquisition; and c) control and shape the liquidity within their exchange by running their own book of business.

Then, once firmly a leader in the exchange market, they acquired the vestigial MoPub business from Twitter for a billion dollars. This was a coup in two ways — it killed off one of the exchange competitors while also increasing the MAX footprint of SDKs among publishers. (As another side note, Twitter’s mismanagement of MoPub was borderline criminal, but I think we can all be thankful that the current ownership does not have that data.)

Overall, a masterful use of M&A to differentiate this company both before and after it’s IPO.

The bull case

The positive story for the company, and one it has certainly leaned into during the meteoric rise in the stock price, is something like:

• We’ve built an AI system for performance at scale.

• We’ve done it in the relatively harsh environment of app installs, and are now taking it to CTV and beyond.

• We don’t need our studio any more — that was a stepping stone — so we’ll become even more efficient.

• Our only real competitors are the big guys like Meta and Google.

That’s a pretty strong story for a public company. It’s also interesting the degree to which the company has benefited from the perception that it has in-app market to itself, since TTD and others don’t really play in app installs. Leading competitors like LiftOff and MoLoCo remain private and certainly do not have the brand awareness or hype that App has achieved. (I would love to hear if people think they get the same performance).

The bear case

Like a poorly-maintained national park rest area, there are a lot of bears. The bears can be categorized into sub-species:

• The “its all fraud” crew. They point out really poor quality ad units running on crappy apps, some of which may have common ownership. Welcome to mobile!

• The “it can’t be real” crew. These are grizzled outcomes-focused marketers who have been there/seen that and just don’t believe the numbers.

• The “cynical ad tech” types (myself included), who are doing pattern matching against previous incarnations of performance phenoms like Rocketfuel and early 2000’s ad networks, who were later found to be generating little actual value.

Maybe we’re all right?

This may be a dissatisfying conclusion, but it may be the case that AppLovin is really light-years better at generating performance for marketers than everyone else (ex-Walled Gardens), and that ALSO, they do this by getting as close to the edge of legality and policy as they possibly can.

Some observations, in no particular order:

• Marketers keep signing up and spending more money. Are they stupid? Do we have any evidence of “smart” marketers turning it off and doing better?

• The various Twitter sleuths have tried to debunk the performance and been mostly unable to come up with any smoking gun of fraud or deception.

• The company very well may be fingerprinting iOS users. So are probably a lot of other companies. It isn’t illegal (in the US). If Apple shuts them down, that will be very interesting, but until they do, why do you care? (unless you’re a privacy nut)

• The company very well might be front-running valuable users to get attribution credit at the expense of Meta. I’m not exactly crying for Meta here. Perhaps this messes up the marketers’ allocations, but so does Google search, and so does Apple’s app store search ads. It’s a complex world!

No matter what happens, this will be an era defining story in the history of digital advertising.

Reading list

• ChatGPT 4o image generation (link)

• This guy Jacob Posel is making ads 24/7 (link)

• Brandtech’s Jellyfish showing dramatic results from AI (link)

• IAB releases publisher server-side thingy and makes a lot of people mad (link)

• Google says news is officially worthless (link)

• YouTube testing dynamically inserted host-read ads (link)

• John Nardone takes CEO of JWConnatix (link)

• 23 & me in bankruptcy, what happens to the data? (link)