As you may know, Apsalar has just introduced uninstall tracking as a standard feature of its Apsalar Attribution platform. App uninstall tracking refers to measuring the rate of installers who go on to uninstall the same app during a specific time period.
It’s important to recognize that no app has an uninstall rate of 0%. Every app suffers at least some uninstalls over time. For more information about the incidence and causes of uninstalls, see this post, or this one.
Using the Apsalar Attribution platform, you can set any length of time you desire, and then examine uninstall rates for different elements of your business. You can do so without paying separately for a standalone app uninstall service because app uninstall tracking is built into our platform and available at no extra cost.
Further, with Apsalar Attribution, you can actually compare uninstall rates for different:
Channels
Campaigns
Sources
Creatives
Cohorts
Regions
And More
Marketers use our mobile app uninstall tracking in order to reallocate spend, collaborate with partners to optimize results, focus investments on high-performing publishers in a mobile ad network, measure ROI for marketing activity of different user segments, and more.
It has quickly become one measure through which marketers can determine how they can drive larger numbers of high quality users.
Given the high level of market interest in uninstall tracking, we thought it would be valuable to provide some use case examples of how a marketer can use uninstalls effectively. We’re going to focus on 6 key challenges:
Campaign Testing
Creative Testing
Partner Optimization
Regional Budget Allocations
User Segmentation and Remarketing
Media Acquisition Model Testing
Let’s examine each of these uninstall tracking use cases in turn.
FIRST THINGS FIRST: HOW TO GET UNINSTALL TRACKING FIGURES IN THE APSALAR PLATFORM
Using Apsalar Uninstall Attribution requires an updated SDK. From there, the user simply inputs two numbers into the Apsalar dashboard:
The Google Project Number
Google Server Key
Then you are ready to go.
Apsalar Uninstall Attribution is fully integrated into the platform, meaning that the measure it isn’t an add-on available only in a single report or solely via a workaround. You can analyze uninstalls in every Traffic and Cohort report we offer. Here’s are two simple screenshots showing comparative uninstall rates for different fictitious media providers.
It’s that simple. Now let’s get started on those use cases.
UNINSTALL TRACKING AND CAMPAIGN TESTING
Testing marketing campaigns has long been a staple of mobile app marketing. By measuring the results across multiple campaigns, marketers can determine which UA and remarketing programs will be most effective, and then adjust their allocations accordingly. Big differences between uninstall rates can be an indicator for why a given marketing effort doesn’t drive more revenue or ROI.
Here’s some example data, pulled out of the platform and reprinted below to simplify our discussion.
CAMPAIGN
INSTALLS
UNINSTALL RATE (7 DAYS)
Campaign A
2,458,621
9.6%
Campaign B
1,984,473
24.1%
Disparities like this can indicate that one campaign will likely be a better performer in both the short- and long-terms. While it always makes sense to focus first on the extent to which your campaign delivers on your core KPIs – like revenue – uninstalls can be a valuable signal, particularly if it usually takes users more than a week or two to drive a first transaction. The uninstall tracking data are only one input in such an analysis, but a valuable one.
Uninstall rates need to be interpreted within the context of scientific validity, meaning that an uninstall rate on a few app installs may not be indicative of what you would see over time. Look for big differences as a directional indicator.
UNINSTALL TRACKING AND CREATIVE TESTING
Just as with campaign testing, creative testing pits one ad or set of ads against another. Creative tests can encompass messaging, offers and aesthetic elements, though the elements most likely to drive uninstall rate differences usually relate to broad messaging concepts and offers.
Binary A/B creative testing is very common in the app world. Your primary focus should be on the extent to which different creative delivers on your KPIs, not (just) acceptable uninstall rates. But uninstalls can be a strong directional indicator.
Many have asked us if we see often see uninstall rate differences by campaign, and the answer is yes. But generally not for small creative differences. Some examples of things that can drive big differences in mobile app uninstall rates include:
Disparate Positionings. Overpromise in creative can lead to uninstalls later, for example.
Offer v. No Offer. Offers can sometime yield far higher conversion rates but then higher uninstalls as well, if people uninstall the app as soon as they have redeemed an offer.
Video v. Static ads. Video provides an opportunity to show in-app experience which sets appropriate user expectations.
Again, it’s important to ensure that you have scientifically valid numbers of app installs/uninstalls for each campaign before you start to draw conclusions.
UNINSTALL TRACKING AND PARTNER OPTIMIZATION
Many clients use uninstall tracking data to compare user quality across different media partners.
Here, marketers are usually looking for big uninstall rate disparities. Big differences gives marketers a reason to dig in and find out why some partners are delivering more high quality users than others.
Naturally, it’s important to put uninstall figures into a broader context. Our primary focus should be on the extent to which a vendor delivers on our KPIs. Uninstalls provide an input in that evaluation, but should not be viewed as the be all and end all. After all, a vendor that drives a far lower CPI but a modestly higher uninstall rate might still be the best value.
Marketers that we have spoken with say they use the data to collaborate with their partners to addresses uninstall challenges. Together, they find ways to mitigate uninstalls so that everyone wins. The client gets a higher quality group of app installs; the vendor a progressively larger block of revenue from that buyer. How can partners use mobile app uninstall tracking insights to optimize? By sharing uninstall rate data with a vendor, you give them inputs they can use to optimize:
The publisher list for your campaign
The ad sizes/format mix
The types of content in which they place your ads
Those are just a couple of examples. But you get the point.
UNINSTALL TRACKING AND REGIONAL/INTERNATIONAL BUDGET ALLOCATIONS
One of the biggest trends we are seeing at the enterprise end of the app publishing market is that companies are taking more and more apps global, introducing them in a variety of regions around the world. After all, in an interconnected world, what’s important is that you have an appropriate phone type, not that you live in one or another place.
Many leading games have been international businesses for years. In other app categories, business unit leaders are only now beginning to globalize their focus.
Uninstalls rates can vary significantly by country. One of the big drivers of these differences is that the leading mobile devices in some fast-growing markets have small memories. For example, an Android phone in India often has only 1-2 gigs of memory, versus 16 gigs for the smallest iPhone. When a phone has a small memory, users are constantly uninstalling apps in order to make room for others.
Whatever the reason for the differences, it’s plain that if a marketer wants to maximize ROI for their app business, they must understand how many people in a market keep the iOS or Android app and routinize use of it in their daily experience. Uninstalls can be a strong directional indicator for the potential lifetime value for users in a given market or region.
UNINSTALL TRACKING AND USER SEGMENTATION/REMARKETING
Remarketing is one of the fastest growing sectors in app media. We’ve seen almost a 1,000% increase in remarketing spend in the past 6 months alone. By analyzing the uninstall rates for different user cohorts, we gain important insights into where to invest remarketing dollars for maximum benefit.
We often see that different user segments have different uninstall rates. Remarketing can help to reduce uninstalls by driving relaunches in the first hours or days of a new user relationship. Personalized push notifications can also make a real contribution.
One of the big reasons why people don’t use an app regularly is that they forget about it. So by concentrating dollars on the users most likely to become regulars, we can maximize our remarketing ROI.
Further, by creating an audience of recent uninstallers against which we can deliver “reinstall” marketing messages and offers, we can sometimes win back people. In addition, we can field research with audiences of uninstallers, to gain insight into the reasons for mobile app uninstalls. The results of such research can help app developers formulate more effective marketing efforts that encompass uninstall marketing.
UNINSTALL TRACKING AND MEDIA ACQUISITION MODEL TESTING
Many brands use a combination of acquisition media models in their mobile app marketing plans. Some dollars are spent on non-incentivized CPI, while other go to preinstall programs, side loading and incentivized download programs that reward a user with game points, virtual goods or free Wi-Fi in change for downloading an app.
App developer marketers usually expect that the number of quality app downloads driven by an incentivized install program will be different than for a nonincentivized one. Uninstall rate comparisons can be a powerful way to discern the extent of that quality difference.
CLOSING THOUGHTS
App uninstall tracking offers a missing link for marketers trying to understand the quality of their users. They are best leveraged as one input for a broad analysis against your KPIs. They can provide powerful answers that marketers need to increase marketing effectiveness and deliver their increasingly steep KPI goals.
Many discussions about uninstalls focus primarily on user experience. User experience plays a major factor in whether or not people hang on to an app. In addition, app store optimization – for both Google Play and the Apple App Store – can also play a role by setting appropriate expectations among potential installers. But there are a multitude of ways that marketers can proactively drive lower uninstall rates just by adjusting their paid marketing channels. By analyzing and optimizing their app install channels. And that is what our Uninstall Attribution is designed to help them do.