Marketing vs Market Engineering

-Free-to-play is a more honest business model than the traditional premium model. And that has consequences for how you market.

-WHAT??

-Let me explain!

With the traditional pay-up-front “Premium” model of selling games, the game company or publisher has an incentive to build up expectations before the launch. They need to make people desire their product. It’s all about the customer expectation. What happens after the customer bought the game is less relevant. If they never even play the game, or test it only once, that’s fine. They already payed for it. I do this myself all the time with Humble Bundle games. I like to think of myself as a person who has the sophisticated taste to enjoy a fine indie game in my spare time. The problem is that I do not have spare time. So I end up buying the games, but never playing them. Which, as I pointed out, is just fine for the indie developers. They got paid for building up the expectations.

Free-to-play is wildly different. The clue is in the name. A good F2P game will let you play the game for free forever. It is better to keep you playing for free than having you not play at all. You might, after all, tell your friends about it, and they might tell a friend who tells a friend who ends up paying. Since you can play for free for several months, before eventually deciding to pay, the players know exactly what they are paying for. And therefore, expectation building ahead of a game launch is much less relevant.

A way to think about F2P is that we are giving the player unlimited try-before-you-buy. We don’t optimise for expectations up front, we optimise for retention of the players who ended up trying our game. If we can keep them, they will slowly spread the game to friends, who spread it to more friends and eventually we end up making money. Have a look at the resulting revenue curves for successful games in both models to see the difference. GTA V and Fallout 4 both made a lot of money in their first week. After that, the sales numbers quickly drop. Candy Crush Saga and Clash of Clans are the opposite. They have made similar money over the games’ lifetimes, but it is much more spread out in time. I looked at the public sales data for GTA V and Candy Crush Saga, and sketched out roughly how their sales were distributed. The curves here show how both GTA V and Candy Crush Saga made about $3 billion in the last 2 years. As far as I can tell from public data, they should be very close to each other in terms of total revenue generated over the last 2 years. That leaves out the original ramp-up of CCS, of course (it was released in April 2012, about 1.5 years before the start of this data series). While the totals are about the same, the distribution in time is very different. GTA has peaks at original launch, and when ported to new platforms (XBox One/PS4, and PC). Candy Crush Saga, on the other hand, is very stable.marketing

If you want the premium curve of high expectations leading to quick revenue at release, you need to do traditional marketing. Creative stuff, buzz building, talking to press. A good game will certainly be needed to convince people that you are worth the buzz, but what’s a good game will be decided by press and early players in a matter of days.

In this mode, you have a lot of activity based on best practises, but they are hard to measure exactly. How much did that magazine interview do to help you? Was it more or less effective than that huge poster at Gamescom? It’s hard to know, but that is marketing. A traditional saying is “I know half my ads aren’t working, but I don’t know which half!”

Free-to-play marketing is different. Just about everything you do is measurable. You have different ads and measure click-through rates on them to optimise. You try different channels and and measure cost-per-install on each. And most of all, it’s not about a huge effort concentrated around the launch date, but about a long-term process that can run over several years. My friend Thorbjörn Warin who has been at Wooga, Grand Cru Games and AdColony had a nice way of saying it. You market premium games, but F2P is not really marketing – it’s market engineering.

If you’re an indie developer working on a new title for Steam, you want to do marketing. The guerilla kind, but anyway. You speak to game media, you’re active on discussion forums, you try to attract Youtubers, etc.

If you’re doing mobile F2P, you might as well not bother with that. Our Benji game gets tens of thousands of downloads every day. A news article about the game usually has such a small effect that we cannot even see it when looking at the download charts. The chart below show tens of millions of downloads for Benji over the last 2 years (the drop in April is just missing data on Appannie). This is market engineering territory.2yearDownloads

Ironically, at the very highest levels of user acquisition budgets, you’re back from market-engineering to marketing. Just look at the TV ads of King, or Supercell’s Super Bowl ad.

At their level, they have maxed out what they can get from ads in other mobile apps. Who hasn’t already seen several ads for the top-3 grossing games (Candy Crush Saga, Clash of Clans, Game of War)? That means that they need to bring in fresh users to the ecosystem rather than take users from other mobile games. But, if you’re the guys making these TV ads, I don’t think you need my advice for how to do user acquisition for mobile games anyway.

Premium on Mobile Today

Premium is making a comeback! Or at least some people hope so..

There are a few companies with an interest in keeping premium alive. That might be good for indie developers.

As everyone knows, the mobile market shifted to free-to-play (F2P) games a few years ago. Back in the Angry Birds era, people seemed happy to pay $0.99 for hours upon hours of entertainment in their phone. Then, a few years later, everyone collectively realised that paying such princely sums for mere entertainment is clearly unreasonable – at least before they thoroughly get to test play the game before being asked to pay. During 2011-2013, the market shifted so quickly that by the end of 2013 between 90% (iOS) and 98% (Google Play) of the revenue was from F2P games.

Since then, there have at times been talk about the comeback of premium. Unfortunately, the data does not support that claim. When looking at the top charts, things have moved even more clearly in the direction of F2P dominance. The current top-100 grossing chart on iOS is 99 F2P games, and Minecraft. It has been pretty consistently like that for the past 2 years. Some hit games that have an existing brand outside of mobile (Console titles, Steam titles, etc.) can briefly make it to the top-100 list – think 5 Nights at Freddy’s or Grand Theft Auto. But the only one that has stayed there is Minecraft.

On Google Play, it’s even more clear. The top grossing game without IAPs today is This War of Mine at position 269. The next is GTA San Andreas at position 363, Geometry Dash at 416 and Star Wars: KOTOR at 431. Of these 4 that are in the top-500, only Geometry Dash was a mobile-first game. That is, one single pure premium mobile first game in the entire top-500 grossing list. the others who got there without using IAPs all had brands from outside mobile.

The best premium games today can make revenue of some millions of dollars. This is for games that get selected by Apple as Game of the Year, such as Monument Valley and Badland. While that is by no means bad, these games also required a lot of talented people to develop. For instance, Monument Valley cost $1.4 million to develop, and generated $5.9 million.

That’s a bit more than 4X return on investment. And that’s the best that you can hope for with this model. In practise, it means that every 4th game that you make needs to be Apple’s Game of the Year.

No investor who can count is going to put their money on such a business. Which means that if your company has VC money, you make F2P games. No discussion! Which in turn might open up some opportunity for indie companies.

You see, there are some companies who would still like to see premium succeed. Apple for one. They are a very premium company and the pay-once-up-front model appeals to them. Premium apps also have the upside that they increase the cost of switching between iOS and Android phones. Your F2P game can be downloaded for free on the other platform, and often you can just keep on playing as if nothing happened. With a premium game, you have to pay again – thus raising the barrier to switching.

Closely allied with Apple in the quest to rescue premium on mobile is the games media. The people who write for games media are often hard core gamers themselves (otherwise, how would they have ended up in that job?). And hard core gamers usually don’t have warm and fuzzy feelings about F2P monetization. In addition, it’s just in their self interest to promote premium games. After all, if I’m required to pay up-front with real money for a game, I want to at least read a review before paying. Which gives games media an audience and a revenue stream. If, on the other hand, I am curious about a F2P game, I will just try out the game instead of reading a review. Goes a long way towards explaining why lots of the top grossing F2P games are completely ignored by the likes of TouchArcade.com. And also why TouchArcade is in financial difficulties as a result of the rise of F2P: https://www.patreon.com/toucharcade

For a large game company, the strategy is clear: develop only F2P games. For a small indie, there might still be a niche for premium. You know you won’t be competing with the big budget companies, and you also know that you have a few good allies if you do premium. A tradeoff of a way smaller best case revenue (by a factor of almost a thousand) versus much less competition could still make it worthwhile. Just remember to keep your budget reasonable. And you might still end up like Zombie Match Defence, if you don’t get a good featuring by Apple.

Or, you might have tremendous success, like the guys who did The Room. They’re another game of the year, though.

dialogue_4

Even more reading about the topic:

http://toucharcade.com/2015/10/16/carter-crater-indiepocalypse-editorial/

http://toucharcade.com/2015/11/13/paid-games-dont-work-for-developers-heres-why-the-carter-crater/

http://www.gamesbrief.com/2015/09/taking-your-f2p-game-and-making-it-paidthe-story-of-jagged-alliance-online/

I’m on CNBC!

CNBC make a short article about the Finnish tech industry – games in particular – and interviewed me for it as well. Over here:

http://www.cnbc.com/2015/11/19/post-nokia-angry-birds-finlands-tech-scene-is-thriving.html

And, I’m pretty sure Tero is ironic in his commentary:

Going Global

Apparently, this blog is attracting quite a global audience. It’s getting read by a few hundred game developers every week quite consistently so far. And have a look at the map – you come here from just about everywhere in the world. Readers in over 40 countries so far!

Thanks and enjoy!

cheers,

Torulf

BlogViews

The Top-100 SDKs you should Integrate into your game

This is about tools and 3rd parties.

There’s a lot of tools, third parties and support companies you need to worry about when building your game. Here’s a brief explanation of who they are and what they do. Unfortunately, a lot of them come with their own SDKs (Software Development Kits) that you need to put into your game.

Let’s start with the platform SDKs. iOS, Android, Windows, Steam and the others will, of course, have their own ways to do lots of things. There are Game Centers of various sorts for social features, billing systems for In App Purchases, and ways to access their unique hardware features such as Apple’s 3D touch on iPhone. A lot of these things are mandatory, and some are just good ideas if you want to be friends with the platform owner – which you likely do.

Next up are Ad Networks. There’s quite a number of these, and they can be used both for advertising your game, as well as advertising other things in your game. That is, you need to integrate the Ad Networks both when you want to spend money on ads and when you want to make money on ads. In both cases, you will want to have several competing Ad Networks integrated.sdk

The problem of multiple Ad Networks has led to two more SDKs that you need. When you are spending money on ads, you want to be able to track how much traffic you are getting through which Ad Network, and how those customers behave once they start playing. For this you need tracking SDKs such as Fiksu, AppsFlyer or Adjust.

If, on the other hand, you want to make money from ads, you have a related problem. Which Ad Network will pay you the most – and be able to fill all ad requests your game can make. For this, you should have a so called “waterfall” through a mediator. You first ask the highest bidder for an ad. If they do not have one, you ask the second highest bidder, then the third, etc. down the line. Who is the highest bidder will vary constantly and depending on what country the player is in. It will really pay off to have someone optimize this for you. Some of the Ad Networks, such as Heyzap and Google, will offer this mediation service in addition to their ads.

Next up, you want to know how your players behave in the game. For this, there are a number of analytics solutions such as Game Analytics, Omniata and DeltaDNA. You can find an overview of them here:

http://www.develop-online.net/special-features/making-games-by-the-numbers-an-insight-into-analytics/0112

Some testing tools, such as the fabulous PlayTestCloud, also have their own SDKs.

Then you will need to integrate towards social networks to get players to invite their friends and thus help make your game more viral. Facebook, Twitter, Instagram and others are in this category. Some also allow your players to record in-game video and post that for others to see. Unity Ads is one of these services. And last, if you find that there are too many social networks to integrate, you might use GetSocial to integrate them all in one bundle.

Some recent addition to the SDK world are the various innovative ways to earn money other than straight IAPs or ads. There are SDKs to let kids ask their parents (or grandparents) for gifts of in-game money. And there are SDKs that let players earn real-world prizes for in-game achievements. There is Creel that links you in-game money to real-world loyalty points (airlines, etc). And there are SDKs that let people bet real money on who is better at mobile games (Gumbler, Skillz, CashPlay, etc.)

Last, you might want to use a publisher to market your game. They also often come with their own SDKs. They might use them for cross promoting between games in their portfolio or for some multiplayer features.

I’m sure I’ve forgotten about some really crucial SDKs here. And rest assured that there are a number of startups inventing new ones as you read this. Just try to think through which ones are really necessary for you, and which you can safely leave out. If you integrate all of them, your game is bloated by all these 3rd party SDKs. You need to constantly upgrade the game just because of changes to SDKs, and your package size might go over the limit of what the App Store allows without demanding a WiFi connection. And if you leave out too many, you might be missing the crucial features that can make your game successful and profitable.

How to increase your ROI by 100,000 X in 5 easy steps

Ad -> AppStore -> Tutorial -> Retention -> Monetization

This post will be about the whole journey of the player, from hearing about your game for the first time, until they are a happily paying customer. There are a hundred nice ways to lose them along the way. The good news is that you can see where they got lost, if you just remember to keep an eye on your players. There’s a funnel from and ad, to the App Store, through the tutorial to retaining the player for a longer time and eventually converting them to a paying customer.

When you are testing out your game, likely in some cheap and cheerful place far away from home (to make sure you’re not measuring your friends), an ad is the first thing the new players are going to see. Advertising is already a science in itself, and you will likely see more than 10x differences between your good ads and your bad ads. I suggest you try out a few different formats on e.g. Facebook and Unity Ads to see what you can get.

At the soft launch stage, when you are only spending a few thousand dollars, you do not need to optimize every last percent out of your ad efficiency, but it is nice if you can spend 1 500 dollars instead of 15 000 to get your first measurable users. On one of our games, we had to pay over 5 dollar per install with one ad, but got under 50 cents with a few tweaks that only took us hours. For one game, Facebook was clearly cheaper, while another got cheaper traffic from Unity Ads. The difference was from 6 dollars, down to 40 cents.

Once you have the ads up, players will be landing on your App Store or Google Play page. I suggest using Google Play during the soft launch phase simply because it’s faster to get new versions of your game published when there is no review process. At this stage, your gameplay video, game description, icon and screenshots will determine whether people who view you Google Play page actually download the game. Happily, Google Play is also trackable nowadays. One early test we did there only got 2% of players who landed on the page to download the game. Another got up to 40%. Again, the difference between a good and a bad version is huge. 20X for us.

Once the players have downloaded the game, they are, of course, greeted by your tutorial and first game session. This too is the subject of it’s own article. Suffice to say, that there is again a more than 10X difference between an OK tutorial and a good one. On our Benji Bananas game, we measured a 40X improvement for the early game retention during the soft launch phase.funnel

When people have cleared the tutorial, we get to the longer term retention. As I explained in an earlier post, this is the most important number that you should track. A bad game has just about 0% 28-day retention. An OK game will likely have a 28-day retention of only a few percent. A really fantastic one can even get above 20%. The difference here is at least 10X, but probably more.

Last, you want to get your players to pay you. That happens only at the very end of this funnel, and with your most ardent fans. Again, you will see a large variation between good and bad games. A badly monetising game will have an Average Revenue Per Daily Active User (ARPDAU) of less than a cent (yes, we’ve been there), while a good one can get above 10 cents. Once more, the difference is more than 10X.

In summary, this allows you to see very clearly why there is a huge difference between the games that make money and the ones that do not. To calculate the return on investment from ad spend to revenue, you need to factor in all parts of the funnel above. If we have five stages, where each can easily vary by 10X-50X, we get a variation of between a hundred thousand and half a billion. Goes a long way to explaining the difference between the top grossing revenues, and the poor indie games that make hardly any revenue. Some indie releases get only a few dollars in revenue over their lifetime, while the top grossing games have generated up to a billion per year for several years.

Go have a look at your funnel, and figure out exactly where you could do better to know what you should improve.

The $100,000 you need to prove that your game works

LTV > CPI

LifeTime Value is larger than Cost Per Install. That’s the golden formula for making money. When you have that, you can spend $3 on advertising to get an install, then wait for the player to spend $4, pocket $1, and spend the remaining $3 to buy another customer. In short, you have a loop that prints money. To prove that you do have that, you will need a minimum of $100,000. Here’s why:

Let’s start with the LTV part. Measuring your life time value is a not trivial. Here’s a short overview of how you can do it: https://deltadna.com/blog/3-ways-calculate-ltv/

The challenge for an indie developer is that you need a significant amount of traffic to be able to measure that number reliably. Say you want a 1 000 paying players to have a good sample. Say you have 3% paying customers among your players (which should be considered quite good). To get a sample of 1 000 payers, you will need 33 thousand downloads for your app (1 000 / 0.03).

The other part of the equation is the cost per install, CPI. This varies according to type of game, and which country you are advertising in. To get good paying customers, you will need to choose a developed country to run your test in. Western Europe, Canada, or Australia might be good candidates, and those countries are expensive. A typical CPI at this point might be $3.

ltv

Now we are ready to sum up: 33 thousand downloads at $3 per download will cost you $100,000. And that’s for a single measurement of LTV with a decent accuracy. Unless you are a F2P monetisation divinity (or very lucky), you will need to do this a few times to get everything optimised.

This is where an indie developer needs a good publisher or a VC investment.

How to get noticed on the App Store

People say that the hardest part is getting your game noticed. They are wrong. That’s part is super easy!

The hard part of game development is making a good game. Once you have done that, the rest is easy. At least comparatively. Now let me start by explaining what a good game is.

A good game is something that keeps players coming back forever. Or at least almost forever. It’s all about the long term retention. The sad thing is that it is really, really hard to build a good game! Once you have done that, all you need to do is get some reliable metrics to prove that you have a good game, and you have a whole bunch of people who are willing to help you in exchange for a ‘modest’ piece of the pie.

Building the game itself will need only developer time (and likely a lot of it). There are no other big investments you need to do up front. Proving the retention numbers is also possible for quite cheap. It is from that point onwards that it will get expensive – or you will need a lot of skill and luck to work around spending real money. I will be exploring both options in coming articles.

If there is a single number that will help you sell your game to partners – be they investors, publishers, the App Store featuring team or others – it is your one month retention rate. Or, to be more precise: your 28-day retention.

First off, what is this numbers and why should we track it?

Your retention number is simply the percentage of users who come back after a specific number of days. For the 28-day number, suppose you got 1 000 downloads on the 1st of February. If 80 of these specific players start your game on the 1st of March, then your 28-day retention is 8%. Players who start the game on the 28th of February or on the 2nd of March do not count towards the 28-day retention (that’s the 27-day and 29-day retentions). Players who downloaded on the 2nd of February also don’t count, of course.

The reason we should track the 28-day retention is because four weeks is long enough to really start measuring if you are able to keep players engaged for a longer term. If you can keep them coming back on day 28, you are likely well on your way to fulfil the 666 rule I wrote about last week. Hardly anyone will be playing a single-trick game without any progression for four weeks straight.

You should track four weeks (28 days) rather than 30 days since there is a clear weekly cycle to retention. Players who download a game on Friday are more likely to play it on Saturday, while players who download on Sunday are less likely to play on Monday. In our games, the retention rate drops by about 5% if you compare the 1-day retention between players who downloaded on Friday vs. players who downloaded on Sunday. That difference can be seen even longer term, which is why 28-day retention is more stable than 30-day retention.

So what’s a solid 28-day retention number? For several years, people were saying that you need to get to 10%. Apparently, game developers are getting better, since the numbers that are now shared by some game companies are likely above 15%, and in some cases clearly above 20%. Of course, the higher it goes, the less risky your game is to commercialise. Which is to say, the higher your retention number, the easier it will be for your to find partners that give you a good deal.

To be able to measure this important number you need to do two things:

  1. Integrate an analytics solution that will let you track the numbers.
  2. Get enough users that are not your friends. You need at least single-digit thousands to be able to measure reliably.

Go here to see a list of analytics tools. I will write more about them in a later post.

To get a few thousand users to measure, I suggest running ads in developing countries. You will not get paying customers this way, but it is way cheaper to measure retention there compared to more affluent markets like Europe and North America. Choose a country where a lot of people speak English for the best effect, such as Malaysia or The Philippines. You will need a budget of a few thousand dollars, but not more than that.

Having a good 28-day retention number does not yet guarantee profitability, but you’re close. Since players stick to your game, you will have time to figure out how to get them to pay. Or you can simply show ads to them.

You also need to be able to show that you can get a decent volume of players into your game. But since acquiring a niche audience is both more complex and more expensive, I will assume you did not do that when you got your test users. Which means that you likely have some mass appeal if you got this far.
During the following weeks, I will be writing about how to track your players from first hearing about your game until they are paying customers. About the all-important retention and its components and how to keep testing and improving your game. After that, we will look into the monetisation bits, why they are much harder to track and what toolboxes you have as a game developer.

Number of the beast, and two times a thousand

The two most important rules we got from analysing the top-100 grossing games.

This time, I am going to tell you the two most obvious rules we got from analysing the top grossing games, as described last week. They are the 666 rule, and the 2×1000 rule. If you didn’t already, I suggest you first read about how we analysed the games.

First off, the 666 rule. According to this, your game pacing should lead to players playing the game:

6 minutes / time,numberofthebeast
6 times / day,
every day for 6 months.

It’s rough, of course. But the point is that play sessions are short, they happen very often, and long-term retention is critical. Let’s take each six in turn, starting with the first: 6 minutes per session.

Short sessions are critical. Mobile games are played throughout the day, whenever the player has even just a few minutes of time. The very shortest sessions can preferably be just a few seconds long – for instance, collecting your gold in Clash of Clans, or spending your gold in AdVenture Capitalist. A longer more involved session might take a few minutes – such as the attacks in Clash of Clans. There’s a reason they are time limited.

The short sessions do not mean that letting people stay and redo the core loop several times is necessarily toxic to your game. It means that one loop around the core needs to be done in minutes. My guess is that this demand is what has killed e.g. the MOBAs on mobile. They have typically demanded longer sessions with deep concentration. That does not work on mobile. I really tried to start playing Vainglory. Several times. Every time, my one-year old interrupted me before I got through the tutorial and the game forced me to start over. After four or five tries, I gave up.

It is also a plus if the concentration needed is not too deep. I regularly see people playing a mobile game while doing something else. My wife plays Match-3 games while watching TV, and one co-workers did an attack in Boom Beach while we discussed new features for our upcoming game. In short, this is very different from PC and console games. They aim for immersion. If you are immersed in a mobile game, you’re about to miss your bus!

6 times a day. That is needed to form a habit. A working game will have you pulling your phone from your pocket every time you have just 30 seconds of time to yourself. There’s always something to do, and every time you pull out the device you make some small progression to your goals in the game. Of course, it does not need to be exactly 6 times. More is better.

Last, but not least, is the aim of keeping your player for 6 months. Or 6 times 6 months, if you can manage. The top games are at the top of the charts mainly because they keep their players for a very long time. High cost-per-install is less of a problem, if the player is loyal for years once they try out your game. I will write a separate piece just about this part – it’s that important.

Then the two thousands:

How will the player progress in game session 1000?
How can a superfan spend $1000?

Calculate six sessions per day for 6 months. That’s about a thousand sessions. Or 100 hours (six minutes per session times 1000 sessions = 6 000 minutes = 100 hours). That’s a nice round number to keep in mind when designing your game play. What will the player be doing during session 1000? How will another six minutes make the player feel progression in the game, after they have already played 100 hours during 1000 sessions? In short, what’s the carrot that gets them coming back at that point? Again, it’s a really hard question to answer, but answer it you must.

This rule of 666 – and the resulting 1000 session rule – was followed by about 95% of the games on the top-100 grossing chart, last time we looked. There is, however, one more rule that got an even higher score. The rule of a thousand dollars.

One of the criteria we looked at when analysing the top games was: is it possible for a superfan of this game to spend $1000 on IAPs? Out of our feature guesses, this was the only one that was present in every single game we looked at. 100%. For some games, $1000 is likely at the upper end. The more casual games make their money off lots of players, each spending somewhat modestly. The more hard-core games, on the other hand, will have much, much higher top spending. Rumour has it that some war games even have oil sheiks spending over $1 million on IAPs. I do not know if it has happened, but at least the games make it possible.

In short, it’s the super fans that pay your salary. Make sure you let them.