Mobile still a moving target
GameChangerSF is finding more value bringing back lapsed players than searching for new ones
The games press, much like the games industry itself, is always forward-looking, always concerned with the "next big thing," the growth opportunities, the potential on the cusp of realization. Looking back through the recent archives, that means plenty of coverage of virtual reality, augmented reality, esports, perhaps some nods to streaming or cryptocurrencies.
Embracing this approach risks neglecting more established parts of the industry, particularly the previous "next big thing," which may still be evolving at a frenetic pace. Phil Shpilberg, president and co-founder of performance marketing firm GameChangerSF, believes that's happening to mobile. User acquisition strategies and tactics have changed dramatically, Shpilberg told GamesIndustry.biz recently, and what worked well just a couple years ago is a recipe for failure today.
"A lot of the changes you see since 2014 aren't necessarily market trends; they're [the industry] saying we don't want to do that; that's not the right way to market these apps"
The company's own data reflects this. GameChangerSF clients represent a wide swath of the market that Shpilberg describes as both too big and too small to handle user acquisition in-house. They handle casual games, mid-core games, and even some of the emerging hardcore mobile market (basically games that aren't social casino titles). Shpilberg said 2014's app installs were commonly driven by incentivized traffic, users compensated for installing specific titles with gift cards or currency in some other game. By paying for incentivized installs, developers could improve their titles' position on the charts, which helped crack the problem of discoverability on app stores flooded with new games.
Fiksu FreeMyApps was the most prominent driver of incentivized traffic on GameChangerSF's network of titles in 2014, accounting for 43% of paid user acquisition on Android and 28% on iOS.
"We hated this during the time," Shpilberg said, "and a lot of the changes you see since 2014 aren't necessarily market trends; they're [the industry] saying we don't want to do that; that's not the right way to market these apps."
Since then, Google and Apple have not only downplayed the importance of the charts as a discovery method, but they've also changed the way they calculate chart placement to take into account user engagement.
"If people were installing then uninstalling your app, you would never make the charts anymore," Shpilberg said. "That went away on Google very quickly. Apple's been battling it more slowly, but the latest iOS 11 basically buries the charts. It was already something you didn't want to do, but now you really don't want to do this strategy."
GameChangerSF's numbers point to a couple strategies developers might want to consider instead. First, the firm has seen an explosion in the proportion of organic installs compared to ones that come from paid user acquisition. Part of that is because of the move away from incentivized install programs, but Shpilberg believes another part is due to maturation of the market.
"The way app stores have evolved over time, and I think this is true of any platform, it starts to favor brands the more the platform ages," Shpilberg said. "Because they have relationships, because the way search works, because the users gravitate to that. It's not like the system is 'rigged.' It's just the way the system shakes out."
"You can get much more scale on Facebook than you ever could before... 70% of our [non-organic] installs are coming from Facebook because they can find valuable users"
In 2014, organic user acquisition accounted for just 16% of the audience on GameChangerSF's Android titles, and 53% on iOS. For 2017, organic installs are 91% of the audience on Android games the firm has worked on, and 87% on iOS. That doesn't necessarily mean paid user acquisition is less important, but it does need to be more targeted.
"We've really moved to a high-quality market where games that make money and monetize their users can afford higher spends and higher cost per install," Shpilberg said. "So with any app these days, you really have to zero in on that audience and understand who those 'good' users are, whatever that means for a game. It could be players who play for a long time, or players who monetize. You really have to go through a lot more to understand them."
Fortunately for Shpilberg and his clients, there are more tools than ever that allow exactly that sort of targeting.
"Before you were dumping all this money into getting to the top of the charts and not tracking what happened after, really," Shpilberg said. "Now every dollar we spend on advertising has to return more than that directly from users spending money."
That's completely reshaped the channels through which games land their customers. GameChangerSF's data suggests there are two particularly significant channels for user acquisition in 2017: the platform holder's own solution (AdWords for Google, Search Ads for Apple) and Facebook.
"You can get much more scale on Facebook than you ever could before," Shpilberg said. "And that's why they're becoming so dominant, at least for us this year. 70% of our [non-organic] installs are coming from Facebook because they can find valuable users."
"Twitter knows what you say, and Google knows what you search, but Facebook knows what you do. And that's pretty powerful"
That 70% figure applies to Android. On Apple, Facebook's share of new users is 71%. AdWords and Search Ads claim 12% and 13% on their respective platforms. In all cases, Shpilberg attributes the success of these channels to their ad optimization campaigns, where firms can target their marketing to an ideal audience - for instance, users with an interest in their game's theme and genre who have a history of monetizing in free-to-play titles.
"People think they're building this giant database of what you do," Shpilberg said. "They are, in a sense, but they don't care about who you are, just what bucket you fall into. So you say, 'I want the 1% of population of the US that look like people who monetized in my game, and they will take the portion of the Facebook population most like the people who did something in your app, and let you target them."
That's not to say Facebook has all developers' needs covered by any stretch. Shpilberg expressed some frustration with the company's offerings, particularly when it came to virtual reality. He said he asked Facebook specifically about targeting VR users, if he could run a campaign for an Oculus game going after just people who owned the headset, and if he could then find out how many of those users installed the game.
"The answer right now is 'no,' because they haven't built that stuff," Shpilberg said. "They don't even see it as a major advertising driver, really, even though they own the company."
It's still early days for the Oculus platform, Shpilberg said, so the strategy for developers is basically to try to get the platform-holder to promote your work. However, whenever Facebook does get around to building out its advertising offerings for VR, Shpilberg isn't likely to bet against the company finding success with it.
"Twitter knows what you say, and Google knows what you search, but Facebook knows what you do," Shpilberg said. "And that's pretty powerful."
He believes it's powerful enough to drive a new strategy for GameChangerSF: retargeting, or going after a game's lapsed players to entice them back into the fold.
"Mobile phone penetration has gotten pretty high, so you're not necessarily getting new people into the market. You're getting a lot of value bringing past users back in with new content"
"Retargeting is where we're seeing some of our best returns," he said. "Part of that is the audience in tier-1 countries is not necessarily growing. For the last two years, mobile phone penetration has gotten pretty high, so you're not necessarily getting new people into the market. You're getting a lot of value bringing past users back in with new content."
One of the most effective ways of retargeting is TV advertising, Shpilberg said, but there are less costly ways to go about it for companies willing to put in some through and work. One game where the company has been trying this retargeting approach is Marvel: Puzzle Quest, which D3Publisher has been running for over four years now. There are some hurdles to this, the first of which is simply identifying lapsed players. Device IDs are one easy way to do it, but it's far from fool-proof given how frequently players can change phones. Facebook's ad optimization campaigns are another method, although they work best when the customers in question actively use Facebook.
Once lapsed players are identified, it becomes a question of how to get them to go back to the game.
"I can't just say, 'Come back to Marvel Puzzle Quest because we love you.' That's not going to resonate," Shpilberg said. "We need to tell you to come back because of these [specific] reasons, like 'We've got these five new characters and we're going to give you one for free.' Then you have to code that in as a deep link and when you start the app, we need to know that's [the offer] you saw and that's what you should get. That conversation with developers is much more difficult because they're fixing bugs and working on new features, then marketing comes to you and says they want you to do this new thing."
It requires a good deal of thought and effort and cooperation from a development team no doubt dealing with its own list of bugs to squash and new features to implement, but retargeting can be a powerful driver for older mobile titles.
"I can tell you that very often, bringing a user back in you know who has spent money before versus getting a new one who you hope will spend money some day, you get a 5x better return on your re-marketing than your user acquisition," Shpilberg said. "But it's not easy; you have to think about it."