Brief History in Mobile Marketing Data Standards

In an earlier post I talked about what incentives need to exist for standardization to work in the mobile marketing industry. I used MATA as one of the many attempts in the industry to break through the standardization problem. In this post I'll go through other examples that my co-founder and I lived through.

As a review, MATA was one of many proposals to standardize the way mobile marketing cost data needed to be organized, calculated, and delivered. Proposed in late 2010 by Kontagent, MATA ultimately tried to make ad networks build their API schemas and reporting in the same way. Great idea in theory, hard to do in practice. It didn't work because ad network incentives didn't align with end users' incentives. Why should the ad network build anything if there is no penalty for not doing it? Advertisers will still spend money on the ad network regardless if their API was built a certain way -- especially if the users advertisers are buying are high quality.

In mid 2011, another standards proposal came across my Tapjoy desk from the attribution providers. The attribution providers' proposal was to pass cost data and properties on the clicks of each of the tracking URLs that app developers/marketers use to track their install base. The attribution companies wanted to take control over the compilation of cost data through the clicks that were getting tracked.

Again, good in theory, difficult in practice. Incentives didn't align. Not only that, the implementation of the technology for passing the data through click URLs yielded additional taxes on all parties.

What were the taxes?

TLDR; In this new proposal by attribution companies, to pass cost data through clicks, the parties were given the ADDITIONAL overheads to make sure reporting metrics aligned on all sides. When there were discrepancies in reporting, the advertiser, ad network, and attribution partners argued for weeks on whose fault the discrepacies were coming from -- this was the "tax".

Let's go into the technical details of the proposal.

When an attribution partner generates a tracking URL, the proposed standard was to pass parameters that can be used to identify a user/device and the cost + cost model about that install. In theory, when this information is passed to the attribution partner, the partner would be able to aggregate the costs from the user level data to the campaign from the ad network.

This sounds good in theory, but in practice, there are a ton of issues why impressions/clicks/installs an attribution provider sees for an ad network isn't the same as the ad network itself. I'm not going to go into all the reasons why impressions/clicks/installs don't equate on the attribution provider and the ad network in this post (I could write a book about it all), but if we take that statement at face value, you can quickly see the overhead that comes with this methodology.

Imagine that the ad network measures 200 clicks, but the attribution partner measures 150 clicks. Simply not having the same number of clicks will mean that the cost data will not line up properly from all the clicks. As a result, the 50 unaccounted clicks, and the resulting costs, will be left up for grabs between the three parties with different incentives: advertisers, attribution partners, and ad networks.

  1. Advertisers will always side with the partner (attribution or ad network) that makes them pay the least.
  2. Attribution partners will blame the ad networks for not doing the calculations properly.
  3. Ad networks will blame the attribution partners for not capturing all the data. Often times they also sit on the short end of the stick because they directly deal with paying the publisher. I've seen ad ops people spend weeks reconciling these issues. Frankly, its not feasible to see a long term solution for them.

As a result, ad networks don't have a big incentive to comply to this standard: 1. for the same reasons MATA wasn't popular -- advertisers still spend money on networks that don't comply and 2. there's a "tax" reconciling data post transfer through the clicks/impressions.

In the end, the incentives didn't align for click standardization to work. Until the end users, ad networks, and attribution companies have the same incentives to change the way things operationally happen, the best way to handle reporting is through transparency.

comments powered by Disqus

Save Money on your Mobile Marketing. Seriously.

Christopher Farm

Love building products people want. Working on tenjin.io - a mobile marketing infrastructure and platform for app developers

San Francisco http://tenjin.io

Subscribe

We'll keep you in the loop with product updates and interesting content around data, APIs and analytics.