The official number that most people cite is $1.80. That’s the average cost to acquire a user in the mobile space, as cited by Fiksu in July. Rumor increasingly has it that the real cost is higher depending on who and where you’re targeting. That may mean $3, maybe even $4, per user in raw advertising spend just to get their install. And who knows whether they are quality users or not.
At those prices the free-to-play model starts to look shaky. It means, for instance, that the player who buys a small pack of coins at $2 (which is the level at which a lot of players only ever transact) is no longer break-even. Not even close when you factor in a platform’s 30 percent cut. While the free-to-play model has often looked to the more serious purchaser to drive profits (the much-sought “whales“), the idea that most initial sales are made at a steep loss in the hope of up-selling customers is pretty scary when 97 percent of players already never pay a dime.
Unlike the sales and star-driven world of movies, television is very meat-and-potatoes. It’s all about numbers-and-traffic for most TV studios, a minefield where pilots come and go, most series live or die with each season, and the instinct of free channels is often to cut and run. There is huge pressure to make mainstream hits that everyone will understand (even though they often don’t work) and massive resistance to anything that looks risky. In games it’s similar.
While the console and PC high-end often seems a lot like movie-land, social/mobile is like TV. You want high retention, high engagement, high lifetime value, high virality and high revenue per daily active user. Set against this you want low user acquisition and development costs. You want content tailored to solve that problem, meaning mainstream and fast-following. That’s why everyone wants endless runners one season, and when they’re perceived as dead, everyone wants the next wave. Ideally what any developer in that mindset wants is to stumble into the next wave idea as cheaply as possible before anyone else, and then be able to exploit it as fast as possible.
It’s all about working very hard to ensure that good numbers stay up and bad numbers stay down on a day-by-day cycle. Keep the inputs low and the outputs high and you’re on the gravy train. This isn’t the business approach that leads to a Year Walk or a Minecraft, but to a Hay Day, a Candy Crush Saga or any one of a number of Bingo games. It tends to attract most investor interest because it’s perceived to potentially be predictable.
However it’s never that simple. In a sense it’s self-defeating because of basic crowded-market effects. Sure, there’s a cool point where it’s all stories like “kid who uses GameMaker to sell 1m copies of a game” or “Game X gained 50m users in a month” but they dry up. In reality some studios get good at figuring out how to leverage one success into another, where others do not. Leader start to emerge and create a downward pressure.
It’s as simple as this: Successful studios either gain a huge amount of quality marketing that makes their game the thing-to-have, or they blast available channels with robo-marketing to the point that their game appears to be the only choice for most users. Occasionally they even do both. Like the TV industry, getting good at stealing the attention of the market in one way or another and then leveraging it is key. You build franchises, networks of players, use one success to fuel another and so it goes.
Unfortunately this process also sucks oxygen out of the market. The skyrocketing cost of user acquisition, for example is largely driven by big competitors trying to maintain chart positions. They are sufficiently financed to drop $100,000 in a day on a campaign and have a big enough audience that the loss-leading $2 users convert into enough whales to justify the expense. They also gain organic growth as a net effect, further ensuring their success.
In the more press-happy end of things the model is not dissimilar but the means are. Rather than naked user acquisition the model tends to become about turning on influencers and telling them a marketing story that they want to relay. The world of previews and promos, interviews and causes often features a comparatively small number of celebrities actively engaged in spreading gospels. It can be hard to get noticed against this wall of folks who already made it. (It must be said that this effect doesn’t really apply in the mobile/social space yet, as there is just generally less of a culture around those markets.)
The sense of a squeeze between the monied and the celebrated tends to become more apparent when a new platform runs out of low-hanging fruit. For example in the touch gaming space there’s been a noticeable decline in startling innovations this year. There’s a sense that touch is somewhat full and the core drawing, tilting and tapping verbs are heavily explored (this is the main reason why I think Apple needs to make iJoypads).
Conversely the squeeze tends to recede when new platforms come to town, when platform amnesia is in the air and existing powers haven’t had the time to establish themselves anew. One of the great aspects of Sony’s push in PlayStation 4, for example, is its drive to get indies onto that system. It won’t last (not to be pessimistic, but historically this tends to be the case) but it is giving some game makers hope.
The reward for those who figure all this out and execute well looks a little like Metcalfe’s Law. The law states that the value of a network relates not to how many nodes are in it, but rather how many connections. A network of 5 nodes has 10 possible connections, for instance, whereas a network of 500 nodes has 124,750 possible connections. It’s an exponential thingy.
A player population is effectively a network with many connections, sometimes looser, others tighter. Its real value is in these connections and whether they lead to strong tribes or flash-in-the pan novelty conversations that fade. As you might expect, building stronger connections tends to be a longer-term and more significant investment like EVE Online. Whereas others look a lot like Draw Something or Dots, or as permanent marketing overload such as all those faceless Bingo clones. Whether mechanical (as in some social games) or cultural (as in gamers), it’s that network that ultimately translates into revenue.
To the outsider success in games looks like magic. To the insider it looks like engineering. Both are half-wrong. It’s never as simple as dreams and talent, but also never as repeatable as features and functions. Talent matters, but so does timing. Data matters, but so does creativity. Resources matter, but so does the platform life cycle. So do the variety of marketing conditions.
That runaway success? Those top-25 developers earning 50% of all App Store revenue? That top 10 on Facebook that doesn’t really move? Mostly they’re examples of studios that managed to sail on a rising tide before everyone else. Whereas the people trying to do it now are in red-ocean territory, churning a lot and watching costs and a lack of press interest drain their investment away.
However the biggest mistake is to conclude that the gaming sector is screwed. If you’d asked anyone what the picture on Facebook would be today given where it was 2 years ago, nobody (myself included) would have said that Zynga would be displaced. It had a 9x advantage over its nearest competitor. Today, however, King is running that table. Likely nobody would dare say at this moment that Supercell or Rovio could be unhorsed given where they are, but again this is not true.
Great gaming companies often rise and seem to dominate, but they rarely do so for more than 3 years. Those that do manage to hang on usually do so because they’re offering a vastly deeper experience, a key technology or have built an unusually loyal tribe (Valve or Blizzard, for example). Mostly what tends to happen is that either platforms get disrupted or pivoted, or successful companies tend to start believing their own mythology a little too strongly and forget how they got to where they were.
Cracks always emerge.
Powered by WPeMatico