Incentivising users with non-monetary rewards to download your app is hardly a new concept. Traditional in-app banners fell out of favor in 2013 when the app store hit the million apps mark and app discovery became extremely difficult. Developers were desperate for new ways to get users, and that's when all the "offer walls" and "app of the day" apps hit it big time, completely changing the market.
The mobile advertising industry has grown ridiculously big since then, with app-install ad spend reaching $3 billion last year. This has brought new challenges. Fraud has become a real problem, representing over 30% of global mobile traffic according to recent studies. With 23% of users only using an app once, user retention is now perhaps the most important mobile metric and advertisers are moving beyond just measuring app installs.
All these changes in the app economy can make you question: how relevant are incentivised installs in 2016?
Brief recap on incentivised installs
Incent ads basically reward the user for completing a specific action, such as installing an app, watching a video or recommending an app to their friends. The user usually gets in-game currency, power-ups and other premium content for completing this task. This is a very quick way to generate installs, build a user base and boost an app's rankings in the app stores. Incent ads are also pretty cheap, as the cost-per-install (CPI) is much lower for incent ads than for non-incent ads.
Non-incent ads on the other hand try to get users to install an app without offering any reward. They work like the traditional in-app banners, only generating users that had some kind of initial interest in the app, meaning that the average life-time value (LTV) of the users is high. Still, there is no guarantee that these users will use the app more than once, and non-incent installs are much more expensive than incent ones.
While incentivised installs can create many users very quickly, there is still debate about their usefulness in the current state of the app industry.
The problem with install campaigns
Incentivised installs are a very popular form of app marketing due to the hyper competitiveness of the app economy. With more than 4 million apps in the major app stores, a new app has a very slim chance of being discovered or getting to the top charts. But with app-install ad spend hitting $3 billion last year, an 80% YoY increase, fraudsters have taken notice.
Recent studies have found that over 30% of global mobile traffic comes from fraudulent sources. Marketers have tried to make scamming app installs harder by using blacklists and paying only for first installs or when the app has been launched multiple times. But criminals keep finding new ways to fake installs due to their high average payoffs, using bots or hiring people to manually install/uninstall apps and costing marketers and developers tens of thousands of dollars. It's even worse for bigger campaigns where over half of the installs can be fraudulent, massively raising the costs of paying per install.
Due to rampant fraud, incent ad campaigns have become less trustworthy. Marketers have started to move away from acquisition of one time users to user retention, which has quickly become to most important mobile metric. Consumer expectations have also changed with the explosion of available apps, leading to 23% of users only using an app once according to Localytics. More and more marketers are focusing on tracking in-app behaviour and retention rates instead of installs.
As the shift towards user retention continues, is there any reason to pay for just installs anymore?
How to use paid installs
The best part of install campaigns is the organic uplift that happens when an app is in the top charts. It basically means that organic, high quality users who haven't seen the incent ads start to discover and download the app. This is especially evident on Android, having the largest smartphone user base but a low contribution to total revenues, where freemium apps do extremely well. The question is how to deal with fraud and low LTVs.
When running an install campaign, it's important to check if the downloads that the publisher provides are actually getting into the App Store and Google Play. The app stores have in-built fraud detection tools and usually reject any installs from duplicate IP addresses, whereas the publishers don't differentiate between these installs.
Checking the log files from trackers and comparing them to the statistics from each app store will give valuable insight on what kind of traffic is coming in and you should obviously not pay for any rejected installs. Some trackers, such as TUNE, also have fraud prevention features that allow you to mitigate fraud by blocking duplicate requests, jailbroken iOS devices and fake in-app purchases.
You can also have higher quality users from install campaigns by optimising your traffic. This is done by targeting your incent ads on multiple offer walls and setting up goals to track besides installs, such as app open rates and session times. Based on these metrics you can check which publishers have the best performance and then give them more traffic by paying a small incent premium, while still keeping a low CPI.
As long as you are prepared to put in extra effort, you can great results with install campaigns. Trackers are becoming better and better at detecting fraud and you can work with the publishers to optimise their traffic to get higher quality users for your CPIs. Discovery continues to be a problem with millions of apps in the app stores, so incentivised installs are still very relevant for the app economy.