closing Game Oven: numbers and struggles

May 27th, 2015

After three-and-a-half years of making weirdly unique social games like Fingle and Bounden, Game Oven closed doors in April 2015. I wrote this article to look at Game Oven’s closure from two perspectives: being a company (the dry bits, Part I) and being a team (the soft bits, Part II). I talk about our studio's costs, our game sales and other incomes, and the role of community to our business, while Bojan and I reflect together on the frustrations we had with each other over the past three-and-a-half years.

In case you didn’t know Game Oven…

Game Oven was founded by Bojan Endrovski (@furiouspixels) and Adriaan de Jongh (@AdriaandeJongh), who were later joined by Eline Muijres (@ElineMuijres). Bojan made Game Oven’s multi-platform game engine (now open sourced) and coded the largest chucks of our games. I (Adriaan) was mainly responsible for game design, prototyping, all business related stuff, a bit of the programming, production, and before Eline joined us, also marketing and administration. With an office in the Dutch Game Garden in Utrecht, the Netherlands, we made Fingle, Planet Challenge (work-for-hire), Bam fu, Friendstrap, Bounden and Jelly Reef.

The Game Oven team.
Game Oven from left to right: Adriaan, Bojan, and Eline.

Game Oven was best known for Fingle, a Twister-like finger rubbing game on the iPad. Fingle was the reason Game Oven started in the first place: we had the game but needed a legal entity to put it on the App Store. (Fun fact: Fingle’s idea came from watching people play the student version of Jelly Reef, which was the project where Bojan and Adriaan met, but which also became the last Game Oven game!)

After Fingle, we made: Planet Challenge (as contracting work), a game to teach kids something about our solar system; Bam fu, a game about punching each other in the face (for real); Friendstrap, a game about being stuck to awkward conversations; Bounden, a mobile dancing game for two people with choreography by the Dutch National Ballet; and Jelly Reef, a game in which you swipe the screen to create a current along your fingers and save jellyfish from harm.

PART I: being a company

Company setup and structure

Game Oven was set up as a BV, which is the Dutch equivalent of an LLC. Setting up a BV is expensive and administration costs along the way are much higher than alternative Dutch structures (such as sole proprietor or VOF), but both our dads had convinced us we could use the experience, so we went along with it.

Our business plan was to make enough games that would collectively bring in enough money to cover the costs of the company, and ‘hope’ for a hit somewhere along the way.

Game Oven employed:

The biggest challenge to our initial designer + developer setup was balancing idea creation / prototyping for games and marketing / doing business for our games. We hired Eline with the promise to overcome that challenge, but when it got real hot in the oven, we still found ourselves helping out with marketing and business to eventually come to face our initial problems: we were perfectly able to balance the development of our games with the time we had to put in other roles, but coming up with game ideas and selling our games seemed impossible to do at the same time. It was always one or the other.

Distributing time between roles with two people.

As an example of a different setup: other Dutch studio Vlambeer clearly separates the producer / marketing / business guy from the game ideas / prototyping / game making guy. The non-overlap in roles allows game ideation to be disconnected from the long stretch that marketing and business can be, so at least one mind can be free and wander around, making weird experiments and test crazy ideas. It seems like prototyping and actual game development are much easier to mix.

On the other hand, Game Oven’s designer + coder setup made up for a super quick iterative process. Our most extreme example was when we were playtesting Bounden every other day, playing a new and improved, sometimes radically different version of the game with external playtesters every time.

Aside from our in-house roles and skills, Game Oven always required at least a number of contractors for every project:

Hiring contractors is more expensive than having people with their skills in-house. Fortunately for us, Bojan comes from Macedonia where the average wage is about one-third of the average Dutch wage. He knew a lot of people there whose portfolio was excellent, and for every game we made (except Friendstrap) we saved a lot of money by hiring someone from Macedonia. Still our contractors were a huge expense to our company, and we naturally asked ourselves why we didn’t just hire people for the skills we needed to make our games. So let’s talk about why we didn’t.

Studio and project costs

One person earning a minimum wage in the Netherlands costs the company around €1800 per month (in Euros, not Dollars!) including taxes and administration fees. With three employees, rent, software, administration, and all sorts of little things, Game Oven cost about €6500 per month.

We think its not cool sharing how much our collaborators asked of us, but the following overview can give you a sense of the costs of each project. Note that usually our collaborators for art worked full time with us while all others did part time:

Costs per project.

* Game Oven didn't pay / didn't exist yet during the first three development months of Fingle.

If you are really paying attention, you would have also noticed that all studio time listed above doesn’t add up to the 3.5 years Game Oven existed. In-between projects, we looked for new concepts, prototyped, maintained the engine & our previous games, and did some contracting work here and there. But more importantly, this in-between ‘creative downtime’ was still ‘studio uptime’ and costs the studio as much as working on a project!

You can imagine what (extra) pressure this put on our creativity. How can you have peace-of-mind when you know you’re burning money every day you don’t find ‘the next big thing’? This is one of the reasons we decided to close Game Oven: being responsible for the upkeep of the studio AND being creative is damn difficult, and we couldn’t find a way.

Studio income

Understanding why and how much Game Oven cost, we can now take a look at Game Oven’s income.

Pie chart of Game Oven's game sale distribution.

Game + platform Release date End date Days Downloads Sales total Average
per day
Fingle iOS 1/12/2012 4/30/2015 1204 269,000 €104,767 €87.02
Fingle Free iOS 12/13/2012 4/30/2015 868 767,000 €12,866 €14.82
Bam fu iOS 5/23/2013 4/30/2015 707 128,000 €3,935 €5.57
Bam fu Free iOS 5/23/2013 4/30/2015 707 266,000 €637 €0.90
Bam fu Android 5/23/2013 4/30/2015 707 1,066,000 €1,330 €1.88
Friendstrap iOS 11/20/2013 2/28/2015 465 8,100 €63 €0.14
Friendstrap Android 11/20/2013 2/28/2015 465 24,500 €90 €0.19
Bounden iOS 5/21/2014 4/30/2015 344 17,500 €26,216 €76.21
Bounden Android 7/3/2014 4/30/2015 301 2,900 €5,527 €18.36
             
App Sales total 11/18/2011 4/30/2015 1259 2,549,000 €155,431 €123.46

Warning: Because my website is SUPER RESPONSIVE and your browser window is not very wide, you're only seeing a selection of the data columns! Check out this article on a big screen to see the entire thing.

Three small notes on the overview above: the sales numbers presented above are excluding the platform's cut, so its the money we actually received. Also, we did not include Jelly Reef as it was released only four days before Game Oven closed doors, and comparing the numbers of a four day old game to the number of much older games didn't make much sense. (For those curious: it made about €6k during it's first two weeks.) And in case you were wondering why Bam fu was downloaded so much: it's free and was heavily featured worldwide in Google Play.

Compared to other apps on mobile stores, these number are not amazing, but they are definitely not bad either. We can only speculate what drove our games' sales, but we figured that a few generalizations about our games would be in place here: Game Oven’s games were unique, potentially appealing to non-gamers, and using phone features in unique ways, but we were constrained by our games working specifically on mobile platforms, which already has a lots of competition, and on which local multiplayer is a niche. We had a preference towards the premium game business model in which every sale comes down to doing marketing well.

(Fun fact: Another constraint to selling our games were Apple's and Google's rules regarding real people or pictures of Apple or Google's devices in the store's screenshots or videos, which was a HUGE pain to the marketing campaign for any of our games. Imagine trying to show someone the fun of Fingle or Bounden or Friendstrap using only screen captures!!)

Something else we found was that many of our players didn't pay for our games.

Pie charts of piracy of Game Oven's games.

Game(s) Downloads
from stores
Unique
users
Piracy Sessions Top
countries
Fingle + Fingle Free* 1,036,0001.04M 1,434,7381.43M 28% 8,426,363 CN (35%) US (11%)
Bam fu iOS 128,000 152,299 16% 726,214 CN (44%) UK (8%)
Bam fu Free iOS 266,000 271,573 - 1,358,214 RU (34%) TR (17%)
Bam fu Android 1,066,0001.07M 1,084,1971.08M - 3,499,791 TR (13%) RU (13%)
Friendstrap iOS 8,100 7,480 - 30,915 US (41%) NL (10%)
Friendstrap Android 24,500 24,761 - 83,916 US (19%) TR (10%)
Bounden iOS 17,500 68,977 75% 173,617 CN (64%) RU (12%)
Bounden Android 2,900 170,223 98% 373,640 CN (80%) US (3%)
           
Total 2,549,0002.5M 3,214,2483.2M 35%* 14,672,670 CN (23%) RU (12%)
TR (6%) US (6%)

Warning: once again, you're only seeing a selection of the data columns! Check out this article on a big screen to see ALL the juice!

Two notes: Fingle and Fingle Free's data are merged because we never implemented separate tracking ID's (and only for Fingle we used Flurry, which can't segment based on game name like Google Analytics can.) Also, the overall piracy percentage of 35% only applies to premium games and doesn't include the downloads of our barely pirated free-to-download games.

Most of our Chinese, Russian and Turkish players pirated our premium games. We can only speculate what drove these illegal downloads. One reason must be the ease at which you can install Android apps outside of official stores. Also, Google Play doesn't sell Bounden in China, so Chinese players looking to play Bounden would have to find it somewhere else anyhow. Another reason could have something to do with payment methods for mobile stores. But most surprising is the amount of illegal installs of Bounden on iOS: China either has a huge jailbreaking community, or something else is happening there that I have no idea of. You tell me.

I think of piracy as an automated marketing tool: many extra people enjoyed our games that didn't necessarily pay for it, but might convince others around them about its value who in turn might pay for it. But another way of looking at piracy is as a missed opportunity to make money. In the end, what we made off our sales wasn’t enough to cover the costs of our studio AND pay our collaborators. We didn't take any anti-piracy measures, but instead hunted for additional sources of income:

Bounden won the Cinekid award for €7,000!
Bounden won the Cinekid award for €7,500!

With these additional sources of income, we were easily able to cover the costs of our studio. We even had enough funds to do another project! Game Oven's income wasn't the main reason we closed the studio, but rather the pressure of our monthly burn rate.

How we think we survived

We believe that a large part of our game sales (but also our other incomes) are due to features by Apple and the game developer community.

Making truly unique games leads to invitations to conferences, festivals and award shows, that lead to making more friends in the games industry, that leads to more friends taking you along to private parties with distribution platforms, that lead to more prominent features in app stores.

If you didn’t think making and having friends around the world is already super fun and rewarding (which it totally is!!!), there is also a ‘commercial’ reason to invest time, money, and friendship into national & international game developer communities. But not in an evil way: you will never have to ask for help if you are willing to help right back. Communities rely on every single individual making investments, including yourself. If you’re not in that kind of community yet, making those investments first could feel scary or like a waste of time, but knowing that people can later make a difference makes every small investment worth it.

Our investments in international friends have been worth it so much up to a point that we’ve stopped caring about the commercial value of our network. It’s like having friends, except it is having friends. Because to have a friend is to be a friend. And there is really, really no evil in that. Members of the game developer community took us to parties, introduced us to other developers or distribution platforms contacts, or gave us advice we just really needed to hear - not because we asked for it, but because we care about those people too. We have spoken out loud about other people’s games because we know how important those games are for them. We playtested, and gave advice. We flew over, hung out, relaxed, and paid for the drinks. We took steps into their direction, and they took steps into ours. It’s beautiful, really, and it doesn’t make it very surprising that so many people want to be part of the international game developer community.

The Indie MEGABOOTH crew at PAX East 2014.
The Indie MEGABOOTH crew at PAX East 2014.

Beside that, our gut feeling tells us that a huge part of the audience for Fingle and Bounden is game developers! Obviously, people close to us and people with similar interests wouldn’t think twice putting down $2 to support us. But beside that small group of people, there is a larger group of developers looking to expand their horizon and definition of what a game is. So letting all our international friends know that our games are out wasn’t just ‘a thing we did for some guaranteed sales’ - no, game developers are our driving marketing force, our community managers! Whenever someone is making a dancing game, developers we’ve reached out to will be the first to tell that someone to check out Bounden! Or when someone asks on Twitter for local multiplayer games, some friends of ours might recommend Fingle!

A critical note on the Dutch game industry

The community I talk about barely exists in the Netherlands. Game Oven made huge investments in Bam fu, and arguably Friendstrap, with false hopes that in our opinion could have been shattered much sooner. A bigger community of more easily approachable, honest, and critical people could have made a difference. We don't blame anyone else but ourselves for making our unsuccessful games, but at the very least a critical community could have set our expectations right and prevent us from investing even more.

For that reason, I announced a game designer meetup called playdev.club last week, where Dutch game designers can come together to play and talk about each others work-in-progress games and prototypes. If you live in the Netherlands or Belgium, don't hesitate to bring your own prototype!

PART II: being a team

As founders, Bojan and I had some struggles during our 3,5 years of working together. They might be presented here as if they were really straightforward, but we went through months of reflection to understand what happened and why we were occasionally frustrated with each other.

Struggle: different ‘deep’ motivations

Adriaan's motivation to make games has always been different from Bojan’s motivation to make games. Adriaan has a thing for truly social games while Bojan doesn’t necessarily. As an example, Bojan really, really didn't like Friendstrap, while Adriaan considers it one of his best working concepts ever. And as Bojan became father after Bounden, money and being sustainable became even more important motivators to him.

Going to the toilet together during a 24 hour livestream of Friendstrap.
Going to the toilet together during a 24 hour livestream of Friendstrap.

Before Jelly Reef, Bojan hadn’t been able to motivate or convince Adriaan to work on any of his game ideas, mainly because Adriaan looked for a social component that Bojan’s ideas never focused on. To Adriaan it felt like Bojan never understood what he wanted to accomplish with his games or that his games didn’t have a bigger vision, and for Bojan it felt like Adriaan never took his ideas seriously. The kind of frustration this caused was very difficult to deal with for both of us.

Struggle: undefined roles

The biggest struggle we faced resulted from never strictly defining our roles. Bojan didn’t want to narrowly define our roles as he wanted to have a say in business and marketing and design as well, on top of being the technical lead - having a say in all aspects of running a game studio was why he wanted to be a founder in the first place. It’s very reasonable, but unfortunately resulted in a few pretty bad scenarios.

Friendstrap

After the underwhelming launch of Bam fu, Adriaan decided that Game Oven’s next production was going to be Friendstrap, which Bojan didn’t feel anything for. As Bojan didn’t have any prototypes for potential projects that Adriaan was motivated to work on, Bojan felt like he had no say in whether Friendstrap should be made or not, but Bojan did feel responsible for it as much as Adriaan did. In hindsight, it is easy to say for Bojan that he was right, that we shouldn’t have spent 2 months making Friendstrap, but back then, Adriaan just took the best decision he could and didn’t see any better alternatives. In the end, defining roles comes down to trust and letting go of things you too feel responsible for.

Bounden’s timer

One long argument we had was about a timer in Bounden. Bojan took on the challenge of making a clear visual representation of a timer to tell players of their progress in a dance. While Bojan was making it, however, Adriaan came to realize that the information that the timer would convey wasn’t as relevant to the broader picture of the game, so Adriaan discarded the idea of the timer while Bojan was getting closer and closer to a good design of a timer. As Adriaan thought the timer no longer fitted in the larger idea of the game, he vetoed it out of the game and replaced it with something more simple conveying different information. At this point, it felt to Bojan as if Adriaan only wanted to include his own designs, while Adriaan was only trying to model a coherent player experience for Bounden.

In hindsight, Adriaan was often not able to communicate his bigger vision fully to Bojan. And what complicated matters were that the execution of that vision changed constantly, sometimes radically.

Bounden’s selection screen

An example of a radical change in execution was the sequence selection screen in Bounden, initially proposed by Bojan. It didn’t fit in Adriaan's larger vision of Bounden and was killed on arrival. But after Adriaan found it necessary to include some sort of selection screen after a couple of playtests, the idea was resurrected. To Adriaan, he resurrect it in a different context with new information he didn’t have before. But to Bojan it felt like Adriaan only wanted to include ideas that he came up with himself. Ugh.

Jelly Reef

Another super difficult scenario was Jelly Reef in its entirety. With our conflicts in mind, Adriaan wanted Bojan to be the creative owner of Jelly Reef. In contrast to previous projects, Adriaan did not intervene at all when Bojan went through getting funding, prototyping, and defining the project’s vision. However, when Bounden was finished and we started the working on Jelly Reef, there was no idea for a structure of the game. And so we went back and forth a lot, with a lot of arguing about what it was that Adriaan wanted out of Jelly Reef and what it was that Bojan wanted out of Jelly Reef. Bojan prototyped visuals and procedural levels while Adriaan worked on a more level design based structure.

We playtested the simple level design based version of Jelly Reef at INDIGO and Adriaan felt he had enough results to continue making the game with a structure that allowed us to include both our ideas AND make the game in 5 months. But to Adriaan's surprise, Bojan didn’t see potential in the version of the game at INDIGO and was confident he could make a fun procedural game. Adriaan advised against making a procedural game at that point, but trusted Bojan he could do it (or allow him to fail). To Adriaan, this felt like Bojan taking complete ownership of the game, which is exactly what Adriaan wanted - but it also meant that Adriaan's role for Jelly Reef became unclear, which wasn’t very motivating and caused Adriaan to be idle and passive for most of the project.

Let’s zoom out a bit here. For Adriaan, not only did he practically leave his previous role as the creative guy behind Game Oven, he also had to work on a game that didn’t align with his deep motivation to make games in the first place: games that help him touch the relationship between him and his friends.

This was when Adriaan announced to the team he wanted to leave Game Oven. Bojan and Eline understood that Adriaan was a fundamental member of Game Oven, so we decided together to finish Jelly Reef and close Game Oven afterwards.

Unfortunately, finishing Jelly Reef with a disbanding team was a nightmare. All three of us had our minds on the future the moment we decided to disband. Eline quickly had a new job where she already spent some time. Adriaan was away for two months showing Bounden at festivals, giving talks, and taking a long planned three-week vacation. Bojan worked his ass off to make it all happen, but was increasingly frustrated with the lack of effort by his team members. It's amazing what Bojan pulled off together with artist Rumena Najchevska and Adriaan regrets he had to bail out on Jelly Reef before it was done.

Struggle: the company’s face

Adriaan was primarily the face of Game Oven. As the more extrovert and better connected, he easily filled in that role. But unfortunately, many interviews with journalist presented him as the sole creator of Game Oven's games even though they were made as a team effort. One Killscreen interview in particularly hurt Bojan most, writing about Adriaan as the genius inventor of Bounden while never mentioning the rest of the team as its creators.

This struggle came back occasionally throughout Game Oven's existence. Everyone wants to be credited for their hard work, and simply making this explicit to journalists keeps everyone on the team happy.

It wasn't all struggles, though!

Only talking about our struggles makes us look like we should have never worked together in the first place, but we obviously had super fun times too!! The many days a lot of stuff got done boosted our asses for weeks, and the team energy was often amazing during those stretches. Taking Adriaan's Unity prototypes to Bojan’s game engine and then quickly iterating on them, building our games step by step: going to the office was a little adventure every morning. We'd make something new, quickly, that day, together, figuring out how the hell to do it.

Even with all the struggles we had, Bojan and Adriaan didn’t close the studio hating or disliking each other. Writing this article was rough on both of us, but we’ve learned from it and can put it behind us to continue to be friends (and have a beer together, soon!).

6 awards and nominations for Fingle and 11 for Bounden, major features in the App Store and Google Play store, stories of relationships crushed and babies made because of Fingle, being on US National television, working together with the Dutch National Ballet, a live performance in the Amsterdam Apple Store, coverage in the New York Times, The Guardian, BBC (reporters in tights = priceless), The Verge, Mashable, and more than 14.5 million play sessions make all of our struggles marginal compared to our achievements. We are super proud and thankful to all our players around the world supporting us and our games.

Adriaan and Bojan with champagne after founding Game Oven on November 18, 2011.
Adriaan and Bojan with champagne after founding Game Oven on November 18, 2011.

Concluding Game Oven

In contrast to what most people initially think, Game Oven did make enough money to sustain itself. Strictly speaking, our game sales weren't enough to keep us running, but only looking at two of our many sources of income would be a narrow point of view. What was difficult for us to maintain was the rigid structure of a studio as opposed to the flexible structure of hopping on and off individual projects. Our problem was 'studio uptime' during 'project downtime,' that monthly burn rate when we didn't have a project to focus on.

We worked hard to finish our games while trying to close the gap between our current project and the next. But leaving responsibility scattered throughout the team by keeping our roles vague magnified the pressure and caused many struggles between us. Eventually, those struggles made us decide to close the studio, and ultimately, that is the more nuanced conclusion to Game Oven.

Our games and the future

Game Oven’s games will still be maintained by both of us and we will continue to work on them together. We have an agreement on how to split the money (50/50), but also on how to decide whether maintaining the app is worth it, and more difficult stuff like that.

Bojan (@furiouspixels) already started working on Plague Inc and more stuff by Ndemic Creations, plus some very small prototypes when time permits. He’s also putting time into being the best dad possible and being outside more often.

Adriaan (@AdriaandeJongh) is currently working on a completely different mobile dancing game and some other funky prototypes (can't wait to start talking about stuff!!!) so subscribe here for alpha / beta testing / releases of Adriaan's games! Also, he's working on a localized version of contract( ) and he became advisor for the Dutch Gamefund!


Comments? Tweet at @AdriaandeJongh.


back to articles, or portfolio