Mobile attribution is the glue holding a fragmented and disparate ecosystem together. Every day, hundreds of millions of dollars are gained and lost between the hordes of mobile advertisers, ad networks and publishers.
And it is all based on the data provided by attribution.
To correctly attribute an install to an ad, several separate parties have to provide unbiased data to the attribution firm. Due to a lack of common industry standards, individual agreements and rules end up creating friction and only introducing further complexity into the equation.
Both Google and Apple block attribution companies from tracking the behavior of users in their respective app marketplaces. The success of mobile advertising rests on knowing whether the user made the download after they saw the ad.
As a result, attribution companies have developed clever ways of figuring out if the user who clicked on the ad is indeed the same user who downloaded and opened the app.
Ad clicks, engagement and impressions within apps, Web and social networks could all contribute to a single install. Attribution firms shoulder the responsibility to keep track of this chaos and ultimately make the judgment about who gets paid.
When the stakes are high–millions of daily downloads and billions in yearly ad spend–precision comes at a premium. Attribution has to continue evolving to achieve the precision that this frenetic industry demands.
A few predictions about how all this might play out in 2016:
Moving Away from Fingerprinting
The two predominant tracking methods – Advertising IDs (GAID for Android and IDFA for iOS) and Referrer ID (only Android) are 100% accurate…when they work. Fingerprinting, the fallback method based on statistical correlation, still has a ways to go.
Mobile operating systems like iOS and Android don’t allow third parties to put tracking scripts into user sessions. Each attribution firm has its own ‘secret sauce’ for determining accuracy of install attribution, but they all acknowledge that they are far from perfect.
Keep your eyes open for an alternative to statistical fingerprinting methods. Ultimately, it’s up to Android and iOS if they want to open up this option for advertisers.
Sensible Payment Distribution
Currently, attribution firms go by a ‘last click’ system to determine which advertiser gets credit for the installation. Regardless of all the ads that cumulatively influenced the user’s decision to make the download, the network that supplied the most recent click will receive 100% of the payout from the advertiser.
Perhaps it should be the first ad click that ought to get the bulk of the payout. Perhaps a 50-25-25 distribution is more sensible. Or how about 40-10-10-40? Maybe impressions delivered on apps with higher LTV users should get more credit? It’s complicated and it’s messy, but some kind of weighted payout is much healthier and more logical than what currently exists.
The ‘last click’ standard is clearly unsophisticated. Changing it would require a great deal of cooperation and trust between all members of the ecosystem. Since change is in everyone’s best interest, we may see movement toward new payment terms in the coming year.
Facebook Taking Over
Facebook created its own mobile app attribution system that runs counter to ‘last click’ and other conventions between attribution firms and ad networks.
Facebook also has data sharing partnerships with a handful of attribution firms that supply the social network with all of their data in exchange for acquisition campaign insights. Facebook in turn feeds on this data to improve its own targeting and attribution methods.
While these insights are valuable for advertisers, and by extension their attribution partners, the data sharing puts more control and information into Facebook’s hands, which may ultimately pull advertisers away from third party advertising networks altogether.
In this industry in particular, knowledge is power.
Due to Facebook’s heft, attribution firms and ad networks can do little but accept these terms. It’s likely that we’ll see further ceding of influence to Facebook as time goes on.
ConclusionMobile attribution is a young, fast-moving sector. As imperfect as it is, the information supplied by attribution firms forms the bedrock of mobile advertising. And yet, despite their importance, attribution firms sit in the shadow of Google, Apple, and Facebook – and any significant change in the industry will require their blessing.