Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
October 25, 2014
arrowPress Releases
October 25, 2014
PR Newswire
View All
View All     Submit Event

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Misadventures in Mobile Advertising
by Paul Johnson on 01/02/14 04:58:00 am   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


Some of you may know that we recently released our first ever free-to-play title, Combat Monsters, on PC and mobile (iOS/Android). It’s been out a couple of months now, and I’ll write a post-mortem later on how that went—in the same vein as all my other PMs; if you can’t wait, the short answer is “reasonably well but could be better.”

But that’s not what I want to talk about now. What I want to do now is to have a really, really good rant--partly to blow off some steam and partly in the hope that mobile advertising publishers may read this and then re-join Planet Earth.

The main reason many indies go the free-to-play route is simply because the lack of a download price removes the barrier to entry and provides squillions of ‘easy to get’ players—at least for a quick look. 

This is definitely true in our case (and I’ve written about this in the past). We have a fairly niche game—meaning there’s no farm (or farm animals of any sort) in it—and there is actual gameplay), so we need to get it in front of a lot of people. The plan being that the small percentage of downloaders who will play a game without the words “clan” or “saga” in the title can still equate to a decent sized audience.

So we started to look at in-game advertising—but the more conversations I had with various providers, the more divorced I felt from reality. I actually started to feel like I was being conned. Seriously, nothing adds up. But first...

Is this actually English?

This is a paragraph of text from the end of a recent conversation in which I questioned a person’s numbers. I’ll get to those numbers in a minute, but I really need to get this alien text down on paper before I explode. Here it is:

Your numbers below are correct, but you're missing a couple variables that should lower that ARPPU number. The primary goal of performance advertising (UA) is obviously to drive downloads directly at a "reasonable cost." But the secondary benefit is that advertising increases the organic lift of your user base. You're absolutely correct that a minority of users will ever pay for an IAP in your games; but the other 95% are absolutely essential for driving more users to your game - whether that be chart position or social sharing with their friends. Further, these "non-paying" users can be monetized if you choose to incorporate ads into your games. Taken together, all of these factors should contribute to an overall CPI goal that backs out in an ROI-positive manner.

Sorry, I just don’t speak this language. Are you using it at me to put me on the back foot? I’m the guy with the pocketbook, so why drive me away with all this technical drivel?

(In case the author recognises this, I apologise for singling it out. All contact I’ve had with advertising providers uses the same insider speak, so am just using yours as boilerplate. In all cases, I explain that I’m a programmer looking to promote my game—so talk to me like one. Please.)

For the sake of other developers who are thinking about advertising their games, I’ve provided a crib sheet of the terms above, based on nothing but intuition.

  • User base:  I think this is “players”
  • ARPPU: The noise my brain makes when trying to figure out buzz-speak
  • Performance advertising (UA): “Performance advertising” does not acronym down to UA. And if there is “performance” advertising, why is there by implication some “non-performance” advertising. Who wants that?
  • Organic lift: Seriously no idea.
  • CPI: I assume this means “cost per install.” But I want to advertise, not install anything
  • Backs out: [Shudders]
  • ROI-positive: Worthwhile?

So let’s look at the numbers. Who wins?

It’s been reported recently that ‘paid acquisition advertising’ (see, I learned a phrase!) was expected to hit $7 per install in the run up to Christmas.

There is an accepted maxim amongst free-to-play developers that you will only convert about 5% of your downloaders into paying customers, for a variety of reasons—very few of which are related to the quality of your game. Combat Monsters is actually double that—but that’s still on message with a generous standard deviation, so I’m going to use that 5% for my illustration.

At 5%, one in 20 downloaders will become a paying customer. At $7 per download, you need each of your paying customers to spend  $7 * 20 = $140 just to stand still. 

Let me just state this again, since it would be easy to not understand this obviously ridiculous proposition: Advertisers expect every single one of your paying customers to be worth $140. Oh wait: Assuming you sell your currency through the App Store or Google Play, you’ll need to actually get $200 out of EACH OF YOUR CUSTOMERS before each store takes their cut and passes back that $140.

Is it just me that finds this preposterous? I mean seriously, what the hell?!? No wonder there’s so much whining about F2P: It may cost hundreds of dollars to play one of these things.

On the other hand, I think it’s clear to see who the winner is...

Do these people live in the same commercial sphere as me?

Several times now, I’ve been contacted by various companies providing some sort of “free app a day” service. The premise for this is that they have 10 bazillion punters checking their twitter feed or webpage every day looking for free games.

I’m usually offered somewhere between 100K – 200K installs from this, but no guarantees of course. And the cost of getting around 150K self-selected freeloaders to try my game?  Oh, somewhere around twenty to thirty thousand dollars. Yes, really! (I typed it longhand so you wouldn’t assume a typo.)

Let’s try a visualization experiment. Imagine you know a small indie who needs help like this—and who just happens to have, say, $25K handy. And from that I can get around 150K installs. Well, we’ve gone past 150K installs on both iOS and Android—and at that point, we’d made less than $25K from either version. And these weren’t people hanging around especially looking for free games either, mostly existing fans of our earlier games and (hopefully) predisposed to try another one.

Tip: The winner is not you... or me.

How fair is this exactly?

Back to paid acquisition: That $7 mentioned above represents a predicted increase from a very solid base of $2-3 that’s usually quoted per install. This still means that you need every single paid user to be worth $86. Outrageous! 

I keep redoing this math because it doesn’t work. There must be a misunderstanding on my part - there must be - but I can’t find it.

Back in the real world, Combat Monsters is currently averaging 13 cents earned per download. We have decent but not crazy numbers to base that on, and a whole bunch of glowing five and four star reviews. Hubris aside, the game is officially “good” according to our players. So 13 cents should be about right then, shouldn’t it?

So why can’t I advertise at a rate where I’m paying at least close to 13 cents per new download?

This is why I’m starting to feel like all of this is a con. There is such a massive disparity between the results I’m seeing—and the assumptions being made about those results by advertisers—that either they’re massively over-valuing their service, or there is something cripplingly wrong with my game’s monetisation... and I really should be a millionaire and not getting stressed by crap like this.

Nope, I really did get it right.

Having developed the pox from the perspective of large online media sites who just won’t cover us, we need to get the word out about Combat Monsters somehow. (We believe the game is far superior to the press-magnet known as HearthStone for example. ) So I went to go set something up as a test. Dip a toe in the water as it were.

Apple’s own in-app advertising seemed like a quick and easy way to try something out cheaply, and I assumed that “it’s Apple” is reason enough to trust that their prices are representative. So off I went to set something up, determined that I’ve misunderstood the true costs I’ve been ranting about, and that all would become clear when I put some real numbers in.

I put some real numbers in. Here they are:

The above is for spending $500 over a week of advertising using iAds. Their own website calculates that from my $500 investment I can expect to see 8 downloads. Really?

We get over 500 downloads a day just from being an available free to try game. Surely they don’t expect me to spend $500 for 8 more?  Who would do that?

Plugging that estimate back in to my own numbers, 8x13 = 104. So for each 500 bucks invested, I get back $1.04. Or put another way, each time I pay Apple $500, I simply lose $498.96. Sounds compelling huh, where’s my check book…

Let me just flip that over and provide another view of the data. Apple expect me to pay $62.50 for every single download they pass my way. They value a download at $62.50, whereas I value it (using actual performance results) at $0.13.

Does ANYONE think this is a good deal? I certainly can’t afford that. I think I need to go find a much cheaper advertising alternative, so will give Saatchi and Saatchi a ring later…

And the winner is…?

(oh, by the way. I’m not singling Apple out here. They’re pretty cheap compared to some of the other things I’ve been hearing. I simply tried them first. Afterwards, there seemed no point trying other sources.)

But wait, there’s more. We sell advert space, too!

Several times now, we’ve read online about developer X who put advertiser Y adverts into his game and instantly made an extra $500 a day. I’m not even going to link to one; the Internet is full of these. 

So we thought, ‘A-ha, those guys not paying any money can pay with time instead and watch an advert’—so we put interstitial adverts into the game for players who’ve yet to spend anything, plus a message that a payment gets rid of them.

We didn’t want to alienate people for nothing, so we tried them at less than full volume in a place that generates 7,000 views per day on average—with 1,000 installs of the app being advertised. In all fairness, this is not a great number—and we could increase it dramatically if we cared to. Here’s why we didn’t.

For those 7,000 views, we get paid about $17. That’s $0.002 (zero point two cents) per view. Or if you ignore the impressions and look just at the installs, each install we generate pays us out the princely sum of $0.017 (one point seven cents) in return. 

Apple want to charge us $62.50 for each download they provide, but this Advertiser (who is meant to be the best payer) is giving us just over a cent for one.

I just deleted another paragraph that was drawing comparisons between those two numbers, but none is really needed is it. Pretty self-explanatory all of this.

And the winner is…?

I’ve finished ranting, so now what?

In 2014, we shall be removing all the various advertising streams from all of our titles—and not replacing them with anything else. That means a couple of our minor games, currently just supported by adverts, will be going totally and utterly free. I really would prefer to just give them away rather than continue being taken advantage of like this. Thanks to the intermediates’ greed, we make bupkis out of them either way.

We’re a small indie firm struggling to pay mediocre wages to an experienced team that deserves far better in my opinion. With that in mind, one thing I will not be doing any longer is making middlemen and sideliners rich whilst we, actual content creators, have to live hand to mouth.

Related Jobs

Digital Extremes
Digital Extremes — London, Ontario, Canada

Sound Designer
Disruptor Beam, Inc.
Disruptor Beam, Inc. — Framingham, Massachusetts, United States

Lead 3D Artist
Red 5 Studios
Red 5 Studios — Orange County, California, United States

Graphics Programmer
Red 5 Studios
Red 5 Studios — Orange County, California, United States

Gameplay Programmer


Paul Johnson
profile image
Someone mailed me this.

You can get a 30 second TV commercial on a minor channel for 50 quid.

Simon Tomlinson
profile image
I was pondering some of these issues a while ago when I wrote my blog on one man app development
_Development_and_the_Art_of_Plumbing.php. I kind of guessed that this was the case; trying to buy installs is really just not economic, but it's interesting to see your actual figures on it. I'm glad now that I didn't stump up what for me would be very large amounts of cash for little real benefit.

The more I have thought about it over the last year the more I am starting to see app development as an internal market. There are ways to make money, but very few of them seem to actually involve making a profit from paying end users of the apps, who clearly will not and do not stump up cash for their endless hours of entertainment. I suspect there is a bubble here, which eventually will burst. The only way I can see to make money without turning completely to the dark side is to do work for hire on apps that are essentially paid promotions or services commissioned by businesses who have no expectation of profiting from the app itself.

Paul Johnson
profile image
Feels that way to me too. As you can see from the piece I linked to about F2P, the numbers just don't work. Not for anybody.

Much is made about the earnings of success stories like Supercell etc., but I have my own theories on the how and why of them affording stuff like this. I won't be buying any shares, let's put it that way.

Christian Nutt
profile image
ROI is return on investment and ARPPU is average revenue per paying user.

The goal of mobile game advertising (as I understand it) is to spend less money acquiring users [= installs] than the LTV (lifetime value) of these users as paying players.

Your ad person is saying that you have to look at the bigger picture and see the value of advertising not just in this binary way (spend vs. how many useful installs you get) but also the overall picture (you need users who don't pay to create more organic growth, i.e. growth not from advertising, but from friends recommending the game to friends, etc.) Basically it's a sales pitch to tell you "hey, ads are more useful than just this one way in which you may be considering them to be." I presume that makes sense as a response to your basic question, right? (You didn't include it.)

I mean, I'm a journalist, not a developer, so my job is to understand things but I'm saved from the trouble of having to DO them, i.e. I don't live or die by this stuff. Some would say, though, that you should understand it better than I do! Because you do! And they would say it in a much harsher way than I just did.

Of course I also suspect you were being sarcastic to a degree for effect, so feel free to ignore this comment if that is the case!

Paul Johnson
profile image
Of course, I do get all that. Originally we only wanted to do some general advertising to build awareness.

(And yes, sarcasm is my thing)

However, since advertising providers display expected immediate returns like this, it's fair to use them to assign a dollar value. And one download is not worth $62.50, even when adding some "organic lift" and other side benefits.

Or if it really is, and this is the real kicker for me, why am I only getting paid under two cents for providing them to others?!? (And not, say, 30 bucks each. I could then use that to offset my own advertising costs.)

But besides all that, if 50,000 impressions typically gets only 8 downloads on average, it means their advertising outlet is ineffective and thus grossly overpriced.

I was initially happy to get 50,000 impressions for $500 until I thought about it a bit more. That extra thinking was brought about by the question to self "how come 50K impressions only gets 8 downloads, I guess nobody is looking?"

Right now, mobile advertising simply isn't fit for purpose. If anyone thinks it is, I'd love to hear their thinking. And by that I mean anyone using it, not selling it. We can all see that it's bloody good news for those selling it!

And just one last point here, using the 5% rule, those 8 downloaders don't yet total a single paying customer. For that I'd need a second and third week of advertising, another $1,000 and a carry flag.

Paul Johnson
profile image
Responding specifically to this:

"Your ad person is saying that you have to look at the bigger picture and see the value of advertising not just in this binary way (spend vs. how many useful installs you get) but also the overall picture (you need users who don't pay to create more organic growth, i.e. growth not from advertising, but from friends recommending the game to friends, etc.)"

This is a point most seem to be missing. This "organic lift" is only a side benefit for this type of advertising. For general products in TV ad's, this type of benefit is all that is on offer - you can't (yet) click the TV to instantly buy a new toaster. So in-app advertising is better than TV commercials and highway billboards.


With TV commercials, your reach is in the millions. And a TV ad has a shitload more "lift" than a crappy little banner ad annoying someone whilst they're trying to play tetris.

This lift from banner ads can only come from those paid for downloaders telling their friends about a cool game they just played. Nobody is going to tell their friends about how they saw a banner ad. In other words, I need to get most of my organic lift from those 8 downloaders I paid for. Not happening is it, might make it to 9?

Machine Works
profile image
Hi Paul,

I understand exactly where you are coming from, it is a harsh reality and the problem is summarized for me in this sentence:
You can't get there from here.

If your game makes you 13c per user/download/day, than you are clearly doing something VERY right.

However, the problem is it takes $$$$$ to get over the hump to make really good money.

If FreeApplesToday or whatever can provide you 150k installs for 25 k , than that is actually a good deal.

Rule number 11: One install gets you a free one - just the k-factor (virality).

So the 24k gives you actually ~300k installs.

300 k installs x .13c = 40k.
However, you don't need to blow all installs in one week - if your game has retention ( e.g, replays)
than you can do 10 x30 k install bursts for a day and actually - just very crudely averaged- 15- 20k
download per day for a month. That's 15k x 30 = 450k downloads x .13c = 58k$-78k$

The next part of the puzzle is that this increases your ad Impressions to the point where can easily do 100k - to 300k impressions per day, with video ads at 10$ per 1000k impressions = another 1500$ per day = 45k - 90k (!)$ a month.
Sum total:
~ 150k. Now you can re-invest in more install, ads, etc.

There are variables, but even if you have to pay 50 k for installs or even 75k for installs,
it is worth doing.

However, you need the initial cash to start the whole chain reaction. And yes, you can find install for 30 -70c.

Again, the problem is you need to get the cash for the PR.
All of this is not fun, but this is how the game is played these days.

All the best,


Paul Johnson
profile image
Thanks Andreas, first positive thing I've heard so far. You'd think even just one of the people I pushed back at over the FAAD thing might have said something like that!

But unfortunately that 13c is (to date) lifetime value, not daily. :( It's an average of total earnings over total downloads.

I now totally get that what I want, I can't have. I'm glad this stuff makes sense to someone, but also depressed that it's yet another avenue that seems completely closed to small companies.

I guess some of this is naievety, but in-app banner ads and the like should not cost more than a TV campaign imo, which I guess is why I've been seeing a lot more of those over the holiday and is now something I will look into. At least I never expect that optiont to be cheap!

Brent Orford
profile image
Making money in mobile is all about statistics. You should be tracking ARPDAU (average revenue per daily active user) and retention % per cohort (the retention % per day of the group of users you 'bought') to really get at the #'s you need. You can't have an estimated LTV for a group of users if you don't know the lifetime. If you track things like retention you might be able to see that paying for 3$ installs will make you more than the .30-.70c ones because they're targeted at people who want to play your game (as opposed to targeted spam) and you'll have a better idea where to put your money to make the most back. Or conversely the small price to spam large numbers of users may yield a higher overall ROI.

Regardless of that, ARPDAU will tell you the most basic thing you need to know. How much money are you getting per-user who logs on per-day. If you have 100,000 users in a day and a 13c ARPDAU... well you'd make 13k - which is pretty good... as long as you keep them engaged and keep stuff available for your users to purchase you'll make the same statistical amount per playing user retained.

Lets say your retention drops off 10k users per day to 0 on day 10, with a start of 100k users and an ARPDAU of 13c. How much money did you make? 13k, 11.7k, 10.4k, 9.1, 7800, 6500, 5200, 3900, 2600, 1300, 0 = $71,500. So your ARPDAU of 13c with a lifetime of 10 days will have a LTV of 0.715c a user. This model's a little harsh, you shouldn't loose 100% of your users in linear fashion over such a short time period but you get the idea. As long as you know your ARPDAU and retention graph, you can estimate how much those users are worth and whether you'll make back your investment in them and then actually make $.

Machine Works
profile image
If your retention is good, you still have a winner.
Good retention (e.g. players coming back over time) is Day 1: 35-45%% Day 7: 20%-35% Day 30: 10%-15%. Retention is the hardest thing to get, period. It means you are on to something and
you can fix the game to make more money. If you retention is not there, build another game or
perform major surgery:).

-Andreas, Machineworks

Paul Johnson
profile image
That's what's killing me. Ours is a niche title, so a lot of drive-by downloaders remove the game almost right away. But those that stay tend to rave about the game, join our forum, spend money, it's all good.

There's just not enough of them. A common problem I know, but one I thought could be dealt with by advertising, given we know we can monetise the right people once we get em. That's not true of a lot games out there, including some of our own, so I'd hoped this would be an option for us.

Our most common first comment is. "Wow, I never knew a game like this existed on mobile." and I thought we could afford to fix that. Apparently we can't.

Alex Feng
profile image
I kind of understand your problem. We are facing the same problem with advertising costing us more than what the game actually earns us. However, my conclusion is that we simply need a better game.

Advertising on mobile generally goes on a bidding basis. If you can pay more, you get more. Channels like iAd is likely not that suitable for games. You might want to look into Facebook or Chartboost. You can bid by either click count or download count. Right now the average bid for per click is about 10 cents and per install 2 dollars. If you don't care about driving large amount of downloads in a short time, you can pay as low as 50 cents per install on chart boost.

What is advised is to pay per install first and see the install rate, then switch to pay per click once you know which group of people are likely to install the game more.

On advertising platforms like Facebook and Chartboost, you can have quite a good focus on who you want to target, so that might suit your need well as well.

At the end of the day, you will still need a great game where players will recommend the game to another to drive down your acquisition cost per player. Adding social features in the game is one great way to do it.

Hope this helps.

Paul Johnson
profile image
Helps an awful lot, thanks. You've already explained it better than the blurb I've been getting from the suppliers. :)

Luis Guimaraes
profile image
"However, my conclusion is that we simply need a better game."

Or a better target audience.

Damir Slogar
profile image
Paul, your math is more or less correct but the starting and ending numbers are not. $7 CPI is ridiculous. Many reports are talking about $7, but this is simply not true. $2 is much more realistic number.
At the same time, LTV of 13c is not good for A F2P game. You need to get to at least 50c to make any UA worthwhile.

Paul Johnson
profile image
Thanks Damir, it's good to have a target.

The good news is our latest number is 15c. Creeping up too slowly for comfort, but at least it's in the right direction.

Tetsu Kamoshima
profile image
Hi, depends on your localisation. Here in Japan the average is $4.5 to $5 or much much more.

Tetsu Kamoshima
profile image
Hi, depends on your localisation. Here in Japan the average is $4.5 to $5 or much much more.

Phil Maxey
profile image
Interesting and entertaining article.

The attitude I have to my current project (iOS F2P fantasy/strategy game) is to concentrate first and foremost on making a compelling game which will keep people playing over a long period of time, and integrate into that situation ways by which the game can make money, but try to do it in a fair way to everyone. In other words monetization is built into the gameplay, but done in a fair (and I think innovative) way.

Paul Johnson
profile image
Heh, you need to read my forthcoming post-mortem for Combat Monsters. What you described is perfectly aligned with our thinking. :)

Jason Long
profile image
From other articles here on Gamasutra, I get the impression that this is actually more of a problem than a solution.

Yes: it absolutely seems like the most ethical and "game-centric" way to go. But like Paul's article highlights: this doesn't necessarily make it worthwhile.

From what I am reading these days, you have to keep your monetization strategy in mind from the beginning, and design the game around it - as opposed to the opposite, which is how I might paraphrase your statement. This isn't to say this method can't be just as ethical or game-centered: it's just not as intuitively so. But in the end, you have a much better shot at success.

I'm reminded of many of my friends who are artists. As you may guess from the common stereotype, many of these people are terrible as business managers. And so even the best artists I know are still poor or floundering - and the end result is that their art doesn't reach many people. Some artists, however - a very tiny few - also have a head for business. And even if their art isn't as great, they are still able to pay the bills and get their art out to more people because they can handle the business side of things.

I'm not trying to imply that anyone is bad at their business: only that what makes intuitively good sense for people who focus on creating compelling experiences doesn't necessarily translate into what is is good for business. And this isn't a bad thing: it just is what it is.

Ron Dippold
profile image
'why drive me away with all this technical drivel?'

I don't think it's actually technical drivel, it's industry specific jargon which has three purposes (for any industry): shorthand for industry insiders, as a means of identifying your fellow industry members (he speaks my drivel!), and actively keeping outsiders (us) out and clueless.

They're not really attempting to enlighten you there; the entire paragraph is a snow job intended to impress you by piling on the cromulent verbiage and leave you feeling that if you don't understand what s/he's saying that you should just leave it to the experts. It certainly sounds impressive, right? Dang, this is deep dark expert only level stuff, right? Of course it breaks down as soon as someone actually tries to pry it apart like you do here. If they had great numbers they'd lay it out as clear as possible. But they can't show you good numbers so they have to run this bluff, which includes the risk of driving you away.

It's very much like the financing snow job you get at a car dealership.

Bevan Davies
profile image
Like others have mentioned, have a look at Facebook for UA, the cost per install is generally better than mobile specific ad networks, especially if you manage the campaign carefully and in general it is targeted advertising so your conversion rate from impression to click & install should be better too.

As for earning money from other people's ads, have a look at PlayHaven and Tapjoy, the amount you earn from other people's ads is going to vary based on your deal with the provider though, but it should be better than what you've seen so far. They also have the added benefit that you can advertise through them as well for reasonable costs if you wish to go further than just Facebook for UA.

Lennard Feddersen
profile image
F2P may not be the right model for all games going forward. Sometimes (and I know this is heresy) gamers may have to pay for certain experiences if they are going to be made. A game you play for hours should be worth at least a Starbucks latte.

Paul, I've started promoting certain kinds of games at Rusty Axe and CM fits the mold. Feel free to shoot me an email and I'll hook you up for a few months. I also have a question about your video that I'd like to talk about offline.

Lastly I'm going to argue that "being a good guy and giving away your back catalog" is only exacerbating the condition. Games cost money to develop and that money has to come from somewhere. If you want to be honest (in terms of the end user experience and how monetisation occurs) and also cut out the middleman then ultimately that money comes from the gamer who is enjoying the product.

Paul Johnson
profile image
Hi Lennard, ltns. Mail incoming, be good to catch up.

Mark Johnson
profile image
Here is how mobile advertising works for us:
Did it again profitably in summer of 2013. We earn much less than we paid for the users we got from the ads. But the users from the ads drive us up the charts where we get many more users for free. Considering all the users we get during the promotion, we earn more than we paid. Its the organic lift. This isn't going to work for every game.

Paul Johnson
profile image
Thanks for this, interesting counterpoint. Although the alternative seems to be "Have mates with massive numbers who will do you a favour". Not sure how sound that is as general advice unless I missed something?