Life Time Value (LTV) must have the highest standard deviation of any of the data-points in mobile performance marketing. For the statistics averse, what this means is that the amount of money each user spends in an app varies tremendously from user to user. Most won’t make a single in-app purchase (IAP) over the course of their journey with an app, even if they engage with it religiously. On the other hand, whales can spend hundred or thousands on IAPs. As Venture Beat pointed out recently, 50% of gaming revenues come from 0.19% of players. That means that for every 500 users acquired, only one will join this coveted cohort. 499 may monetize by viewing ads, a few may make a small purchase or two, but ultimately your average revenue per user (ARPU) is being propped up on the shoulders of rare giants.
The obvious question is: Why then are advertisers paying the same amount for each user when they are far from a homogenous pool of spenders? The short answer is that media buyers can’t yet target at the user-level even with the most advanced audience segmentation strategies we currently employ. The best we can do is target cohorts, users with a certain set of common data-points, but even still the combination of possible cohorts is practically infinite. The task of pricing them individually would be massive and wouldn’t be worth the trouble.
There are indications however that whale hunting will become a more transparent and achievable goal. Some channels like Facebook and Google may have the ability to target on a user level because of the massive datasets and personal information they guard. As well, targeting based on device ID or behavioural metrics can be useful in accounting for audience members’ mobile patterns. Still, user level granularity is far off and nothing can guarantee that an ad will be seen by a whale. At present, we can’t hunt whales with a single rod in the ocean, but we can cast a wide net in the right waters.
Of course, ‘the right waters’ is a vague description. Where are these waters? How much of the ocean do they constitute? Effectively, we are talking about the location/timing/messaging of ad-placements used in audience-cohort generation. Creating cohorts based on metrics like publisher ID, Creative ID, GEO, time-stamp etc. allows agencies like Curate to zone in on the places whales frequent, on what messaging they respond to, when they surface, and the like. This means we can markedly increase the probability that we are reaching them. Advertisers are still not able to purchase on the user-level with these methods, though they can tier CPI rates based on e.g. which publishers, creatives, and GEOs perform well
Moving forward, the notion that as long as CPI < LTV the advertiser has profited will slowly yield to a more granular process by which advertisers use engagement metrics to indicate what user spend will be. These Cost Per Engagement (CPE) campaigns do a better job of matching ARPU to ad spend, but still don’t account for the majority of users who will engage without generating revenue (CPM monetization notwithstanding).
We’ll be publishing a blog series over the next few months to address this very issue and explore strategies being used by mobile advertisers and agencies. Stay Tuned!