menu close
menu close
back to Insight Page

TIL: Facebook’s Retention Optimization

Written by: Erik Mansur, VP, Product and Engineering

This is the latest post in our “TIL (today I learned)” blog series that puts a spotlight on specific digital advertising technologies and strategies. Today we give you the lowdown on Facebook’s “Retention Optimization.”

WTF is Retention Optimization?

Retention Optimization is another effective way for mobile app advertisers to use Facebook’s vast data to refine their acquisition strategies.

It joins other Facebook ad targeting methodologies App Event Optimization (seeks new installers likely to complete an in-app action — usually a purchase) and Value Optimization (seeks users likely to drive downstream revenue). At its core, Retention Optimization helps advertisers acquire users who have shown a willingness to return to a mobile app in the days after downloading it.

What problems does it solve?

Mobile games can be divided into two categories: those that earn revenue via in-app purchases, and those that do so through in-app advertising.

If you’re an avid mobile gamer, you’ve probably played games with time-based challenges for which you’ve willingly — or begrudgingly — purchased items within the game (like coins) to help forward your progress. Facebook is aware of this purchase activity and how frequently and how much you spend. It then helps advertisers use that data to put more personalized advertising in front of users, through the use of Value Optimization and its cousin Minimum ROAS bidding.

A 2015 study by Microsoft made the claim that the attention span of the average Canadian(!) was now less than a goldfish’s. Mobile devices have arguably helped cause this change in behavior. The frenetic way that users interact with devices has encouraged game studios to offer experiences that require less time and focus.

Related post: TIL: Facebook’s Minimum ROAS Bidding

Hence, the new genre of “hyper-casual” mobile games that are easy to learn and allow for quick, pick-up-and-put-down engagement with the app. But the studios that create these games still need to generate revenue, so their hyper-casual games often include interstitial ads (videos, playables, etc.) that users are forced to view at certain break-points during the game.

Because most hyper-casual games lack a clear-cut way to monetize — either via an in-app “store” or a freemium “get-rid-of-the-ads” model — Facebook’s Retention Optimization is an ideal fit. It allows these companies to seek users who will achieve a lower-funnel goal that supports a revenue model based on user loyalty and advertising.

While Value Optimization and Minimum ROAS Bidding work great for games with in-app stores, they don’t make much sense for advertising-based hyper-casual games. Why optimize toward in-app buyers or the minimum value that can be driven from those purchases when you have nothing to sell them directly? Conversely, optimizing towards upper-funnel objectives like installs or registrations may help drive new downloads, but that often leads to paying for a lot of installers who play the game once and never return, which limits the amount of in-app advertising that can be displayed.

“Because most hyper-casual games lack a clear-cut way to monetize, Facebook’s Retention Optimization is an ideal fit. It allows these companies to seek users who will achieve a lower-funnel goal that supports a revenue model based on user loyalty and advertising.”

What’s most important for hyper-casual gaming companies is to find users who will end up seeing more in-app ads, and Retention Optimization is the best way to get there.

What are the best ways for me to use Retention Optimization to make my advertising campaigns successful?

One of the advantages of this new methodology is that it comes in two types. Advertisers can optimize for users likely to return to mobile apps within two days or within seven days of installing.

If the game’s back-end metrics shows high engagement from early user interaction — and therefore more in-app revenue from advertising — then two-day Retention Optimization is the way to go. If the game provides less-intrusive advertising  (i.e. interstitials delivered at a lower frequency), then seven-day Retention Optimization may be your best bet.

Value Optimization allows advertisers to refine what “value” actually means to them through the use of Minimum ROAS Bidding. Retention Optimization offers something similar, but using a different, yet still familiar, path: Lowest Cost with Bid Cap bidding. While most advertisers will begin testing Retention Optimization just using Lowest Cost bidding (f.k.a. Facebook “Autobid”), implementing a Bid Cap allows them to specify the precise cost ceiling they’d be willing to spend to acquire a loyal user.

While this post has covered Retention Optimization through the prism of a particular game developer, it’s important to remember what this technology actually does: help advertisers find users who will revisit their mobile apps more frequently within a given timeframe soon after installing. So non-gaming app advertisers could also use this type of optimization.

Related post: TIL: Facebook’s Dynamic Language Optimization

For example: an ecommerce app with an immersive user experience — plus data showing that more user interactions correlates to more purchases — could use Retention Optimization to target installers who return to apps more frequently, which may turn them into buyers more quickly.

However, the most obvious use case for Retention Optimization is the hyper-casual game that depends on in-app advertising, and there’s no reason for these game developers not to test this new technology now. It’s rare that an advertiser can say this, but Facebook has created an optimization type that’s hand-tailored to revenue goals. It’s also being released at the perfect time of year: the holidays. Millions of users will receive new mobile devices as gifts. They’ll have time to play newly downloaded games … and be exposed to the advertising inside.

next post

But wait, there's more

Join Our Newsletter