Blog Archives

Chasing the (CPI) Dragon; the ever-increasing cost of mobile user acquisition


Will Freeman has written a good article for PocketGamer that examines the increase in mobile game user acquisition costs over the festive period.

I find this upward trend and one-dimensional thinking to be very sad and worrying for what it means for the larger mobile games ecosystem. The ‘cost’ is not simply fiscal (e.g. in terms of CPI) but much wider in terms of the impact it is having on the wider mobile games ecosystem. I believe that it is damaging to games makers and failing games players alike.

I recently attended the App Marketing Summit in Berlin and many speakers there were proudly talking about how enlightened developers should focus on profitable installs (where LTV>CPI) instead of simply aiming for high chart position or ‘simply’ building huge DAU/MAU levels. Whilst this is true, it is also obvious (sell stuff for more than it costs to make and market….duh!) and doesn’t address the fact that we are fast reaching (or have reached?) the point where only well-funded studios with a portfolio of products, strong IP and which already have an existing audience, have a chance of being commercially sustainable. Is this not the exact same scenario that drove developers away from the traditional triple-A console/PC games market?

My point, however, is not that the temporary utopia where anyone can make a mobile game has been eroded, because it is still perfectly viable for anyone to make a game on mobile. Rather, what vexes me is that if you make a GOOD game that is polished, technically robust, novel, great fun and demonstrably ‘worthy’ of an audience (e.g. the metrics show people engage with it), without the aforementioned commercial capacity and advantages you will have zero (OK, maybe 0.1%) chance of a major breakthrough. There is very little chance, even, of achieving financial breakeven. Ergo, mobile game development is not commercially viable for most game developers.

When the entire industry focuses on using paid-for user acquisition (in-game adverts and cross-promotion) then it is only focussing on ‘pushing’ to that vast ocean of mobile device users, and pushing is proving to necessitate an ever-increasing amount of marketing spondoolies. Focussing only on ‘beating’ the CPI is, ultimately, a battle we are destined to lose as CPI is only ever going to increase. Paying for users is an addiction that this industry has and, currently, accepts but one that needs to overcome for our collective health.

Paid-user acquisition is also highly geared towards supporting large-scale F2P games that successfully milk a minority of users or premium games based around very strong IP that people will pay for before properly trying the product. I have no axe to grind with either – I’m a big fan of F2P, when done well, and who doesn’t find a Minion funny? – but it makes the range of games available very limited in nature. It leads to an endless number of ever-so-slightly improved clones of a small range of game genres and themes (fancy another match3 or endless runner game anyone?).

That means that the vast potential for smaller-scale but creatively and technically innovative games, or simply games aimed at smaller niche audiences, is completely neglected. This isn’t a plea for artificially propping up wannabe indie dreamers; this is about wanting to engender a genuine ecosystem where consumers (our users, our customers, the nice people what pay our bills) get the games they want and those that make them can afford to do so. Should it not be possible to make and games for, say, £100k and to achieve net sales (after Apple, hosting etc) of £200k? Would that not be a good commercial outcome? Is that not profitable? Could that be sustainable?

The paid-user acquisition model is ever so wasteful; it is like paying a bunch of rich foreign fishing boat owners to hang a 10-mile long fishing net off the back of dozens of boats to catch one million sardines when you only want (and can only sell!) the couple of hundred tuna that also get caught. Everything else gets thrown back into the sea. The rich boat owners are the plethora of advertising and cross-promotion networks and the dead sardines get resold again and again to the next unsuspecting and naive customer (aka games developer). Now you can argue (as indeed I have done) that this ‘waste’ is irrelevant so long as the net (pardon the pun) result is that there is more cash in your bank account that when you started. How sustainable is that though? No matter how clever you are with data analysis, ultimately games making in involves creative and commercial risk. I keep hearing that F2P is now well-understood, yet we all see plenty of F2P games continue to fail commercially even though they, presumably, survived the ‘culling of the ugly children’ stage of development when the metrics are not strong enough. The costs of production and the cost of marketing are rising sharply. Would not a more intelligent approach be to try and find ways to acquire users at a lower cost? This industry has always been at the forefront of technology development. It is strange that it isn’t using that expertise to improve the ways in which we match games with gamers.

Paid UA will always exist and will rightly always be part of the marketing mix but right now it is pretty much the only strategy and vast sums of money that are being spent in/on games is leaching away from those that make the games to those that operate the advertising networks. Does it really make sense that those who spend months or even years toiling away, taking creative, technical and commercial risks are simply the means by which another industry (advertising) gets wealthy?

Don’t hold out for the app stores of traditional games media to be the panacea that fixes this problem. The app stores are the means of distribution for all mobile games but the means of discovery for only a tiny proportion of games. This will not change to any significant degree. The vast, broad mobile games audience does not engage with traditional games media. The dozens of ‘discovery apps’ that are out there currently are (largely) characterised by paid-for promotions and, quite often, dirty underhand tactics to promote shoddy games or to grab advertising revenues from unsuspecting users.

I am convinced that we need to be thinking about how we can ‘pull’ users into games by understanding the interests, behaviours, actions and characteristics of individual users and applying all kinds of (well understood and proven) predictive modelling and targeted recommendation techniques to show them the games that they will more likely enjoy and more likely play for extended periods.

That is what we are working towards; “Making it easy to find good mobile games you will enjoy”. If you are open-minded (or just plain utterly desperate) about the state of mobile games user acquisition, then please drop us a line. We’re largely in stealth mode still but keen to talk to anyone in the industry about how things can improve….hell, even fat fishing boat owners.

Register your interest at or mail us at


Make your game dynamic


Knowing where your new players came from (which user acquisition source) and knowing something about them can be a powerful way to maximise that first session experience in order to maximise initial retention.

SWRVE is a platform that allows you to perform split testing (A/B testing) in social/mobile games and they provide regular webinars on all manner of subjects relating to games analytics. Here’s my summary of their advice in relation to adapting your game based on where the new players come from.

The main acquisition sources are likely to consist of ‘organic’ (social/viral), incentivised installs, promoted adverts/media and cross-promotions. Each of these tell you, generally speaking, that those cohorts of users have a propensity to respond more or less to different elements in game and in different ways.

Players that were acquired through genuine organic (FREE!) sources e.g. through discovery via the Facebook newsfeed or notifications or by a direct message from a friend are most likely drawn in by the social interaction aspect. In this case, SWRVE argue, you should ensure that any social elements in your game are exposed to these players very promptly. This might include opportunities to visit friends’ worlds, donate resources, gift in-game items etc or it might be to challenge them in, for example, a social quiz game. Make sure that these features are available and promoted in game early on to leverage this player behaviour preference.

Players that are acquired through incentivised means (e.g. when they receive free resources in one game in return for installing another game) are commonly thought of as ‘poor quality’ traffic and used most often as a means to get AppStore visibility in order to then acquire better users in a more organic fashion. SWRVE argue that you should still aim to maximise retention from these sources by, for example, ensuring that the core value of your game is exposed to these players with the maximum of haste; let them know why your game is fun immediately. Additionally as you know that these players respond to incentives; offer them rewards straight away (earlier than other players).

Players that are acquired through non-incentivised adverts/promotions in other games clearly have the propensity to chug out of one game in favour of another. To that end, showing them in-game ads to 3rd party games is not a great idea unless you like losing the players you just paid $5 CPIs to acquire.  Make sure that in-game ads and cross-promotions are switched off in the first few session to these players; give them a chance to get engages…or better still, to start paying. If after multiple sessions these players show no sign of converting to becoming a paid player then you can switch those promotions back on in order to monetise the traffic from them.

As a subset of the above group, where players have come to you via a video acquisition source then you can assume (again, generally) that they have a preference to video-based media….so use video in the game to keep them engaged. This might be for tutorial elements, scene-setters or for moments of delight when they level up or defeat an opponent.

As an addition to the above, where users have connected via the Facebook API and where you have added demographic information about them, then you can use this to further make your game personalised to individual players in order to improve your KPIs. Examples may be with in-game mentor character choices/genders, colour schemes or language style).

Naturally, SWRVE advise that you don’t just make sweeping generalisations when electing to adopt any of these strategies; make sure you test alternative strategies on statistically identical groups and prove which approaches deliver the optimum KPI improvements. Then ramp up across your entire user base.

You can see all of the previous SWRVE webinar recordings online at

%d bloggers like this: