Blog

Share
Publishers

In-app header bidding: The good, the bad, and the reality

by Sean Webster on May 15, 2018

Leveling the playing field for publishers is a great objective, but header bidding may not be a good fit for in-app advertising.

Throughout the short but active history of programmatic advertising, there have been many initiatives to make the auction process more equitable for publishers that too often find themselves surviving on too small of a piece of the ad revenue pie. Will the latest hyped method—in-app header bidding—end up falling into the same category? Let’s break the model down a bit and explore the realistic potential.

The rallying cry behind the in-app header bidding movement is familiar: “Let’s level the playing field, make every auction a real one, and make the whole process transparent.” It’s a noble idea, but execution is a challenge. To borrow a line from Felix The, writing for VentureBeat, header bidding is “a very smart hack,” but the technology is simply not ideally suited to in-app advertising.  

In-app advertising demands multiple technical adaptations to implement header bidding and requires ad networks controlling the market to make room for another party, thereby willingly foregoing revenue potential, which seems unlikely to be viewed as reasonable by the networks and mediators.

But if you can set these significant technical and business challenges aside for a moment, it is easy to be intrigued by the buzz around in-app header bidding.

The good

In the desktop scenario, header bidding helps publishers earn more by letting them monetize each and every impression for its maximum value via “…a fair, transparent, simultaneous auction where the only factor determining the winning bid is the bid itself,” according to The Drum.

When it works, header bidding, in theory, is a square deal for everyone:

  • Buyers get the impression they want at the price they are willing to pay;
  • Publishers are rewarded with more revenue for every sold impression; and
  • Users get quality content for little or no cost.

The challenge in applying the header bidding model elsewhere is that in-app programmatic—as we’ve all learned by now—is nothing like desktop programmatic.

The ecosystems aren’t the same, the technology isn’t the same, and the majority of app publishers simply don’t have the business structure in place to make in-app header bidding work.

The bad

Header bidding is less effective for in-app advertising than in online ads for multiple reasons.

First and foremost, the ad ecosystem for in-app ads is much more complex. Referring again to The’s article for VentureBeat, header bidding “…is a hacky workaround subject to the feature limitations of each individual ad server. The ad server space in mobile in-app is far more fragmented than web, and each ad server has its own set of auction mechanics and rules.”

In other words, while the online programmatic market is mature enough to have distinct players with defined roles in the food chain, in-app programmatic is a whole different world with buyers and sellers at opposite ends of the spectrum with often incompatible goals, intermediated by a single (yet variably modeled) ad network/mediator that’s free to do things as they see fit.

Secondly, while applying header bidding techniques in the online world is a relative snap (just drop a bit of code into your page headers and watch the market open), creating the same capability in-app is a difficult task involving the application of engineering resources to enable SDK updates or adapters.

That’s more expensive than online header bidding, it takes longer to do, and it requires ongoing maintenance. So, even if the revenue from in-app header bidding was equal to the online method, ROI would still be diminished due to higher initial and ongoing costs.

Finally, while online publishers of scale—and a small handful of the larger global publishers—have dedicated sales teams to drum up interest for their inventory in global markets, the vast majority of app developers and publishers have no such capacity. That said, it is extremely difficult for most app publishers to develop the relationships required to maximize the potential of header bidding in foreign markets.   

The reality

In-app header bidding can look good from afar for cash-strapped publishers, but upon closer inspection, this new programmatic technique may be far from good for many.

In some ways, the whole idea of in-app header bidding is that it’s really a solution to to automate a process that currently exists. In today’s mobile programmatic landscape, various ad networks like MoPub and AdMob (and of course AppLovin) already allow advertisers to appear in multiple instances within the programmatic waterfall.  

While the concept of header bidding is appealing (i.e. instantaneous auctions open to all demand sources, including exchanges, networks, and direct buyers), the reality is that, technical restrictions aside, in order to work for in-app, those who sit at the center of the programmatic market (mediators and networks), would probably have to choose to side with either buyers or sellers and allow the entry of a new player to represent the other half of the equation.

That will be a difficult pill to swallow for networks and will ultimately add more mouths to feed in the programmatic market, meaning a reduction in opportunity for publishers—exactly the problem header bidding is supposed to solve in the first place.

Add it all up and the projection is that much of the promise of in-app header bidding may remain unrealized.

Sean Webster is AppLovin’s senior director of Business Development.

We’re hiring! Apply here.