What’s in store for app marketing in 2018

0
SHARE


Editor’s note: The following is a guest post from Alex Tarrand, Lucktastic’s vice president of marketing.

If 2017 was anything, it was educational. Social and machine learning tactics moved into the political theater, major gaming companies pursued vertical integration and some bought their best user acquisition sources. Apple also graced us with yet another App Store redesign. We’ve been through a lot, and 2018 will likely be no exception. Here are some predictions for 2018 as we skid further into the new year.

Old media, new distribution

Los Angeles has cupcake ATMs, and they’re as amazing as they sound. Everyone loves cupcakes, and giving people a new way to buy them was a master stroke. Similarly, some game and app genres are always going to kill it. “Clash of Clans” made strategy massively multiplayer online video games casual and Sony is now augmenting every “Jeopardy!” episode with additional questions on Alexa. Game shows are moving to the web, a medium where modern audiences spend the majority of their time. This type of content will always be popular, but the nature of distribution is shifting. More traditional media companies will look to platforms like Facebook Live or Instagram Stories to broaden their audience.

Asking more of acquisition

Most companies have shifted to a “return on advertising spend” (ROAS) model. What this means is that marketers are ditching cost per install (CPI) and only paying once a new user performs a qualifying action, such as initiating game play or leveling up. Given the ways that fraud continues to evolve, advertisers will be looking for more qualifications that a new user is going to be valuable and loyal.

The rise of probabilistic modeling

Most attribution partners will provide a built-in long-term value (LTV) predictor. However, as acquisition costs continue to rise, marketers will start to build probabilistic models that will granularly predict a player’s value within a day or two. High-level models are blunt instruments, meaning you need to dive into your app’s specific metrics to really predict lagging metrics like LTV. Using basic machine learning business intelligence teams can output models that will help predict players’ value as well as prevent players on the fence from leaving. Services like Amazon Web Services machine learning make formerly complex correlation analysis affordable and relatively quick. When you have a solid LTV predictor, your team can make timely affiliate optimizations based on that data.

Apps break into TV advertising

When I first saw SuperCell buying a major ad spot during the super bowl XLVIII, my head exploded. My assumption has always been that it was less of an attempt for an ROI and more of a nod to fans. Up until now, TV ads for mobile apps used to stand out. But over the last year, it’s become commonplace. That said, plenty of companies still don’t have confidence that they will see a solid return by running a primetime spot. The good news is they don’t have to. There are a wave of remaining TV agencies that are de-risking TV buying by making it more financially accessible, and they’re finding better ways to quantify spending. All this is making TV more competitive, and marketers will look to generate a profitable CPI, regardless of their customer LTV.

Go direct or go home

Attentive managers will corner you in bars and tell you horror stories about finding that perfect site ID that just can’t send them enough traffic. So, why can’t that sub-source send you the millions of installs that you would kill for? Because they’re probably selling a good chunk of their inventory directly to savvy partners. We’re going to see more and more companies seeking out direct deals with their best sources to lock in strong traffic with good margins.



Source link

LEAVE A REPLY