A Developers’ Guide to App Analytics. Part 4: Revenue Attribution

Your app downloads are up. Great! It must have been all that marketing you did. But which campaign worked best? Which partner should you re-hire? For the answers to these questions, you need attribution analytics. 

“The best things in life are free. But you can give them to the birds and bees. I want money. That’s what I want.”

It’s a 60-year-old song, but its theme is universal. Money. It’s what we all want. OK, so there are some app developers with higher goals than profit. Maybe you have an ethical mission, for example. But even high-minded developers need income to keep their objectives on track.

So it’s vitally important to identify the factors that turn your users into spenders – or ad clickers.

In a series of posts on this site, we have explored the world of app analytics. We looked at tools that show you how your users are behaving inside your app. Then we examined the tools that feedback the key numbers – downloads, retention churn, session length, and more.

In this last article, we will dig into the world of revenue attribution. 

Today’s app developer has so many options when it comes to marketing. The in-app advertising has exploded – it was worth $6.7 billion last year

There are hundreds of ad partners promising to display your ad in just the right place at just the right time to drive downloads. You can run campaigns in other apps, on desktop or mobile websites, or on social channels. And you can measure results in different ways too: clicks, impressions, installs, events, payments.

Revenue attribution tools will help you to navigate this complex landscape – and make informed decisions about which strategies to repeat, and which to abandon. 

As such, your attribution partner can help you measure the following key data points:

Organic installs
You need to know how many users downloaded your app organically, in order to separate this group from those who responded to some kind of incentive or ad.

Paid installs
These are installs coming from active promotions – advertising campaigns, incentivized downloads, etc.

Post-install activity
What does the user do post-download? Activities can include registration, ad clicks,  adding items to cart, making an in-app purchase, buying a subscription, and more. Your analytics partner will help you link these activities with the initial source of the download.

Mean time to install (MTTI)/ mean time to action (MTTA)
These data points reveal how long it took for users to go from click to install or from click to some other conversion point. 

View-through attribution 
Sometimes users might see an ad for your app and respond by downloading it – without actually clicking on the ad. View-through attribution lets you see where this might have happened.

Average Revenue Per User (ARPU)
This is the average for all users.

Average Revenue Per Paying User (ARPPU)
This is the average for the subset of users who have completed a payment within the app.

Lifetime Value (LTV)
While ARPU measures historic events, LTV is a predictive measurement. With this model, you can make informed decisions about the worth of your app/company and your future marketing activity.

Return On Investment (ROI)/Return on Advertising Spend (ROAS)
Obviously important. ROI is the barometer of your marketing activity, enabling you to compare one partner with the next.

Having access to the above metrics is useful, and your analytics tool will let you peruse the results on your dashboard. But the real value comes from linking outcomes to a specific activity or third party.

For this reason, attribution specialists have links to the many thousands of adtech firms, and to network/channels such as Facebook, Google, Twitter, Snapchat, Pinterest, etc. 

This means you can select your desired partners on your dashboard in order to correlate an app action only to them. 

Exactly how analytics firms link actions with partners is complicated. And it’s getting trickier. This is due to privacy concerns. Historically, tracking companies have relied on unique identifiers buried inside a phone to do their work. But earlier this year, Apple announced it would scrap its IDFA identifier. Google hasn’t confirmed what it will do with its GAID identifier for Android. But industry insiders think it will take some kind of restrictive action soon. 

For this reason, analytics companies are exploring other ways to make accurate measurements. We will look into this in a future article.

Linking an action to a specific partner is extremely useful. However, sometimes you’ll find that your user has interacted with multiple partners/promotions before doing something that resulted in revenue. 

This raises an obvious question? Which source should take the credit?

It’s another area in which a good analytics company can add value. It can show you which source delivered the first and last ‘touch’ for example. It can even offer ‘fractional attribution’, so you can award a percentage of the rev share to multiple partners. 

As with all other analytics tools, you can set useful filters that make your metrics more insightful. These include all the expected options (handset, location, time, etc). One especially useful filter is the ‘lookback window’. This defines how far back you want to go from the time of install/action. The ultimate goal here is to make your campaigns as effective as possible. 

Most tools have a seven-day lookback default, but this can be modified. In general, the longer the lookback window, the higher the conversion rate since users have more time to convert. But on the flip side, this approach can over-inflate your numbers and attribute more credit than campaigns deserve. 

While analytics tools can ultimately help you see what’s working best to earn you revenue, they can also be useful in identifying what’s ‘too good to be true’. Sadly, ad fraud is a fact of life these days. Unscrupulous actors have devised all manner of cheats to manufacture clicks or take credit for others’ hard work. They even create device farms that perform repeat actions – such as registrations, installs, and engagement. Advertisers like you payout, but there are no real users. A good tracking tool should keep you alert to these dangers.

Ultimately what app tracking tools do is help you understand your customers better. In the analog world, this was always so difficult. But with connected products like apps, you can know where your customers came from, how long they spend in your app, what they like best, and what turns them off. There’s almost no limit to what you can study. 

Over the last four posts, we have given you a little snapshot of what’s possible. Now it’s over to you.