MoPub’s Optimizer Lets Mobile Publishers Automatically Prioritize Their Most Lucrative Ad Networks

MoPub is releasing a new tool called the Optimizer that should allow mobile publishers to take an entirely automated, hands-off approach to managing their ad networks — and increase revenue, too.

The team gave me a demo of the new feature, saying the technology uses a “waterfall” approach, moving down a list of possible networks from which to serve an ad, starting with the one that had the highest estimated CPM (price paid per thousand impressions). Normally, MoPub prioritizes those networks based on CPM estimates provided by the publisher. The problem: Those estimates are often wrong. (MoPub has been trying to address the lack of transparency and data about the performance of individual ad networks with its new dashboard.)

Now, when publishers hit the Optimizer button, MoPub will automate that prioritization process based on its own data and the data it has acquired from various networks, so that it can predict the likely CPM, clickthrough rate, latency, and more on a given ad. Ideally, for each impression MoPub should be serving an ad from the network that’s likely to make the most money for the publisher.

“It seems like a really simple concept, but our publishers haven’t seen anything like it before and they’re basically blown away,” said Marketing Director Elain Szu.

The data used for prioritization is supposed to be as specific as possible. In other words, when possible, MoPub will calculate CPMs and so forth using data specific to that publisher and that geography, but when necessary it will use more general data, and in some cases, when there’s really no data available, it may just fall back on the estimates provided by the publishers.

Not every publisher is going to embrace this for all of their campaigns, Szu added. Instead, she suggested it could be particularly useful for small publishers who don’t have the resources to manage their ad networks in a very hands-on way, as well as for larger publishers who may have a number of geographic segments to monitor. Those larger publishers may want to pay close attention in more mature markets like the United States while taking a more automated approach in small-but-growing markets.

The MoPub team also showed me the results of some early campaigns, particularly how the share of ads from different networks shifted when the Optimizer was turned on, and continued shifting over time. (In some cases the Optimizer would even shift money away from the MoPub Marketplace to other ad networks.) In each case, the CPMs went up compared to past performance and compared to apps that weren’t using the Optimizer — you can see one example in the (anonymized) chart above.