Ad Tech Do-Over: Header Solutions & Monetization
August 25, 2017 - Originally published by
The benefits of header bidding on the web are well documented: Publishers receive an average increase of 15% to 30% – even as high as 50% – in CPM, advertisers win the impressions that mean the most to them and people have more relevant ad experiences.
On the web, adoption has been swift: Header bidding exploded on the scene about three years ago, and since then more than 70% of top publishers in North America have adopted it.
The next step is to bring bidding to apps to balance value for advertisers while maximizing monetization for publishers.
App publishers tend to be more technically focused, swift and incredibly adept at monetizing their properties. So, why is bidding in apps almost nonexistent?
One answer could be that advertising is a smaller percentage of revenue for app publishers. The biggest players make much of their money through in-app purchases and subscriptions. Since advertising is only now being added as a second revenue stream, app advertising is in its infancy.
Or perhaps it’s even simpler: There's no “header” in apps.
It's more technically challenging and time-consuming to create a “wrapper” – or, more precisely, a bidding clearinghouse – for an app. You have to ensure the server-side code works seamlessly with the bidding SDK. Publishers would have to add another SDK, which many are loath to do.
With load times already slow, another SDK – or another three SDKs – is a hard case to make. Therefore, the tech will need to be server-side, which adds to the sophistication requirements for direct bidding in apps.
Furthermore, the tech ecosystem for apps is underdeveloped and fragmented across both ad serving and mediation. Current solutions are still based on the same decade-old waterfall that often biases their own demand. Best I can tell, only Amazon Publisher Services has developed third-party tech that is even close to bringing direct bidding to app. Early indications are that MoPub has desires to move in this direction, but its timeframe is unclear.
So, what will it take to move the market?
First, the stack can be simplified. There are only three necessary functions in today's world:
- Planning and trafficking
- Serving ads
- A bidding clearinghouse where all demand sources can compete equally, with the highest price winning
(Quality – human x viewable x transparent x filters – could be considered a fourth pillar, but for the purposes of this discussion, I'm sticking strictly with the supply “tech.” I'll save quality for another column.)
The first two functions are already progressing as the natural market forces seem to be working. Nativo, Freewheel and others have evolved the ad server and with that, the incumbents have felt the pressure and are evolving, too. So, for functions No. 1 and No. 2, check the box.
On the third layer in the stack, header bidding was a hack on ad servers on desktop, proving that a fair clearinghouse makes more money for publishers. So where is the bidding clearinghouse layer in apps? Time will tell whether Amazon, MoPub or someone else will lead.
There's even an opportunity for the ecosystem to create an open-sourced version. In web, there were two widely used libraries: AppNexus’ Prebid.js and Yieldbot's Pubfood. For it to be viable in app, publishers would have to be willing to contribute. Is anyone out there willing to contribute? Only time will tell.
Unfortunately, publishers are faced with the paradox of choice. Currently, no mature bidding clearinghouse exists. So, what to do? Do nothing or work with waterfalls until something better comes along?
Both options lead to an inevitable repeat of past ad tech failures. Quickly the waterfall tech will bias its own demand, waive fees and provide volume incentives. Pure tech players who are trying to offer a fair and transparent option aren't given a chance, incumbents take back control and publishers are disadvantaged yet again.
Mobile app is a chance for the ad tech do-over. Let's propose a different path for us all: Refuse to work with waterfall vendors, build your own solution or test anyone willing to meet your requirements.
The time is now.