AdExchanger caught up with Radic to talk about Tapstream’s tech.
AdExchanger: What problems are you trying to solve at Tapstream?
SLAVEN RADIC: We’re about bridging the gap and eliminating the friction between mobile sites and apps and between apps and ad clicks. For example, Twitch [acquired by Amazon in August], uses us to help move people from a web session into the Twitch app so they can seamlessly continue their experience.
It’s a common problem that apps have to deal with. Let’s say you’re on the Yummly mobile site looking at a recipe and an ad pops up that says, "Why not try our app for a better experience?" So you do. You download the app – and suddenly you’ve lost what you were doing on the site and you need to navigate from scratch. The app treats you like you’re a brand new customer. Yummly is a client of ours and they actually had this problem.
So, how do you make the experience better?
We’re able to advertise with more specific calls to action and landing pages. Most people would bail if they were dumped on the home page of an app, but if there’s a connection between a marketing click and an app so that the app responds with an intermediary landing page suited to the marketing offer or that simply continues the user experience from the website, that experience will be much more effective.
What about user acquisition?
We’ve been in the attribution for almost three years. We started by integrating with hundreds of ad networks. In that way, we’re similar to Kochava or MobileAppTracking. We have an SDK in a customer’s app that triggers and sends data back.
But we also have some new things rolling out, one called “onboarding links” and another tool called “word of mouth,” which allows us to bring push and fingerprinting to new levels of accuracy.
You can have great creative, but if you have a bland onboarding experience that doesn’t relate to the original message in your ad, the app is going to have an insanely high abandonment rate. It would be like if eBay or Amazon didn’t use landing pages on their websites. With the onboarding links, we’re helping app marketers pay more attention and get more aggressive with their marketing.
Our word of mouth is an app referral program, which personalizes content automatically and rewards users for the installs they help generate.
How do you strike the right balance between focusing on installs and user lifetime value? Installs are important, of course, but they’re nearly meaningless if an app can’t retain its users.
With the right data and the right strategy, you can win at both. The Zombify app is an elegant example of this. If you align your data with your users, not only will you drive installs, you’ll also drive engagement.
People like to tell their friends about the apps they try, but posting on Facebook or tweeting isn’t as direct as being able to send your friend a personalized text message about it, which you can do through our word-of-mouth tool. That person’s friend learns about the app, and the referrer gets to unlock content for sharing – which ticks the install and engagement boxes. We want to get our customers thinking more deeply about acquisition, not just marketing spend.
You used the word “fingerprinting” before. What do you mean by that?
Fingerprinting, the way we define it, is a much more benign form of tracking than any other. It’s ephemeral and it doesn’t last beyond 24 hours. It’s actually the best case scenario for consumers, because no IDs are exchanged and nothing is stored about a specific person – no demographic information at all. When we talk about fingerprinting, we’re talking about attribution.
Your onboarding links feel a bit like personalized deep links.
In a sense. We’re an attribution company at our core, so we have analytics in the back end to connect clicks to installs and to track engagement. Our client Yummly, for example, saw that they were 50% more likely to retain a user if that person came through an onboarding link.
What data do you collect?
Further along in the process, our SDK also collects data about behavior, including whether a user made any in-app purchases, registered an account, sent a link to a friend, tweeted about the app, etc. Those are all things we can measure and use for optimization.
How do you use the data?
We track a variety of variables, but we don’t track people. We take a temporary fingerprint which is only relevant for a short period of time, and we store this bundle of information on the back end. When an app runs our SDK, it’s one of the first things that fires, which sends more information to the server, where it’s matched with other data.
We don’t use IDFA or advertising IDs to make connections across platforms. We do temporary fingerprinting using our own homegrown algorithm to create matches in the back end. Then we coordinate the data to the marketing channel, because certain channels are better for certain things. One channel might be better for installs or registrations, while another is better for encouraging purchases.
What data points are app developers and advertisers most interested in?
Recently, this has changed completely. Last year, people only seemed to care about cost per install and the number of overall installs, but today it’s about engagement. Revenue is actually a lagging indicator of success. For most apps, engagement is everything.
It’s becoming much harder to get a person into an app because there’s so much competition, which is why once you’ve got there there, you need to engage them and keep them. Install is only the first battle.