Understanding attribution: The economics of mobile marketing

Rising to the top of the App Store charts isn’t just a matter of buying enough installs. Developing a profitable marketing strategy that keeps spend low and ROI in the black requires planning, forethought, and hard data delivered by a marketing attribution provider. The data from an attribution partner provides the necessary insight into a user’s entire journey through your app, going beyond analytics to give a more detailed picture of how they arrived, how they behave when using an app, and how those choices vary across different cohorts.

Today we’re taking a look at the the economics of mobile marketing. Who does the buying and the selling of mobile ads? How are prices structured? And how do you structure your KPIs to align with your marketing budget? We’ll take you through the basics. If you’re new to marketing attribution and want to learn right from the start, check out our beginner’s guide on what is attribution.

Who buys and sells mobile ads?

There are different players involved in mobile advertising, chiefly buyers and sellers. On the buying side, there are app developers. They’re the advertisers - they want to get the message out about their brand to the world.

On the selling side, there are publishers. Publishers provide and sell the space to run ads. If an app advertises in different locations but also sells ad space within its own app, it is both a buyer and a seller!

In between the app developers and the publishers are ad networks. Networks are the go-between; they connect the supply of ad space to the demand of the advertisers. Sometimes app developers also work with agencies to run their marketing campaigns.

The following chart shows how those different entities are positioned within the ecosystem, and where your attribution provider should lie in relation to advertisers, publishers, networks and agencies.

An attribution provider like Adjust is an unbiased third party, whose job is to attribute credit to an ad campaign source or sources. This can be a challenge if a user looks at or clicks on multiple ads in their buying journey. Our job is to find the particular data point that’s responsible for the action they took. We are trusted by both advertisers and publishers to attribute data points and resolve discrepancies when they arise. By positioning ourselves separately from the parties are either selling or buying mobile ads, we’re able to maintain a position of neutrality and offer analytics and data that is objective and reliable.

How are ads bought and sold?

Between 2010 and 2015, time spent on mobile grew a mammoth 700 percent. 90 percent of the time spent on mobile phones is in-app. Global mobile ad spend topped $100bn in 2016. In short, mobile advertising is a massive market - yet only 49 percent of marketers have a mobile analytics solution to help them figure out if they’re actually earning any return on their investment.

For those with an attribution provider in place, there are four major cost models to apply when working with ad networks. The model an app chooses is based on their KPIs, which usually differ by vertical (meaning that gaming, e-commerce and travel apps all have different goals - this determines how they set up campaigns with ad networks).

• CPI (cost per install): This is the most common way to buy or sell media. An app will pay a set amount of money for every new install. The cost can depend on the app and the estimated lifetime value of the new user. The network receives part of the CPI price for finding the best publishers for the specific ad campaign, and the publisher receives the rest.
• CPC (cost per click): In this model, the advertiser pays a price for every single ad click. These are not as popular with advertisers as CPI campaigns, as the cost of an install can increase exponentially with all of the ad clicks that happen. Also, advertisers must rely on networks to tell them the ‘accurate’ number of clicks that happened, which can represent a conflict of interest.
• CPA (cost per action): These campaigns charge advertisers after the user completes a designated task. It could be signing up for the app, playing the first turn in a game, or buying a subscription. Action-based campaigns are useful for advertisers keen to see a return on their investment.
• CPM (cost per mille): A ‘mille’ (French for 1,000) refers to 1,000 impressions, or 2,000 eyeballs that have seen your ad. If a campaign is, for example, geared towards a brand and there is no specific call to action (to download an app or subscribe, etc), it is typical to use this model. As with CPC pricing, the advertiser relies on the network to tell them how many impressions took place.

How does paid traffic affect my organic traffic?

A user clicks on an ad, gets taken to your page in the App Store, and downloads your app. What happens next? Well, a few different things: that user might tell people about your app by word of mouth. They might post about it on social media or refer a friend through your referral program. With enough installs, your app might start to climb the App Store rankings and make it onto one of the top charts. From there, more and more users will discover your app organically. These users will continue to discover your app even as you discontinue your user acquisition campaigns. You can expect organic installs to increase as you increase in the App Store rankings - this is what’s known as organic uplift.

You can calculate your organic uplift - here’s an example. Perhaps you have an app and you don’t buy any installs, so you receive thirty in a single day. On the next day, you buy two hundred installs, and you watch as your organic installs increase to seventy over the course of the day, up from thirty. The next day you buy five hundred installs and you receive two hundred organic ones. On the day you bought two hundred installs, your organic uplift is your seventy organic installs minus the thirty organic installs you would have had with no UA campaign, divided by the number of paid installs, or two hundred. That’s an organic uplift of twenty percent; on the day when you bought five hundred installs and received two hundred organic ones, your organic uplift is thirty-five percent.

A huge burst of paid user acquisition campaigns just after launch can set a strong trend of organic uplift in motion, enough to keep you in the App Store charts long enough to grow your audience to the size you need, which is vital if your app has a community aspect to it, for example a game with forums or a dating app.

What kind of marketing budget do you need when launching an app?

In a recent webinar with Grégoire Mercier, CEO at Addict Mobile, the question of how much a user acquisition campaign costs came up. Many people don’t know what to expect when they’re embarking on their first app launch, Mercier told us. You have to make some decisions up front - how long will the campaign go on for? Will you go local or go global? Will you have a soft launch first?