Nathan Lovato

Nathan is a content crafter and Game Design & Analytics expert at GameAnalytics. When he's not doing research, you can find him writing and pulling colorful worlds out of his imagination.

We got in touch Melanie Christin, the co-founder of Atelier 801, an independent game studio that produced the massively multiplayer Free to Play title Transformice. Its community is extremely active since 2011, year of its initial release. We asked her for tips and feedback on the company’s experience.

Melanie, can you tell me who you are and what you do at Atelier 801?

Hi, I am Melanie Christin, the cofounder of atelier 801. We built this game company with my partner, Jean-Baptiste Le Marchand, thanks to the success of our first game, a multiplayer title called Transformice. And we created the game one year before the studio.

Your flagship game is Transformice. How does it play?

Transformice is a massively online multiplayer browser-based game. The player is a small mouse that jumps from platform to platform to catch a piece of cheese. But he’s competing with other mice to be the first to bring the cheese back to the hole. One of the players is the shaman. He has the power to create crates and planks to help all the others in their mission.

This is our main product today. It represents 99% of our revenue.

Playing Transformice is a cheerful and funny experience

The game is made in Flash. Why did you go with that technology?

First of all, Transformice was released back in 2010. Then, both Jean-Baptiste and I were working at Ankama on the MMORPG Dofus. He was a developer there and I worked as an artist. We both had a lot of experience with Flash. Especially Jean-Baptiste, who had already created many games with that technology: he was one of the few persons back then still specialized in this language. Thanks to his experience, we could reuse a lot of code to make Transformice. It took us less than 3 weeks.

Flash has other advantages. As we use vector graphics, the game was and still is lightweight today, after 7 years. It’s but a 1 MB download.

Only 3 weeks? Did you plan to do Free to Play from the very beginning?

No, not at all. We made the game in 3 weeks. It was just a basic version, a test project for us. We released it on a forum,  for free. We had no intention to monetize it. But suddenly, we had 10 000 players and an overloaded server. People had to queue up to play. We needed to pay for dedicated servers, which are very expensive. That’s when we decided to put an Adsense banner below the game, on the webpage. It generated enough revenue for us to keep the game running.

One year later, we left Ankama and created our own game studio: Atelier 801. Back then, will still relied on ads exclusively. However, it was enough for the both of us to make a living. it is only 2 years after we released the game, in 2012, that we decided to transition towards Free to Play.

How did it happen?

Well, that wasn’t planned either. At the end of 2011, we were really happy: we had a fresh company, our first employee… We had huge activity spikes, millions of players, all was good. But one day, we got banned from Google AdSense. And as you can imagine, we didn’t know why and had no way to get in touch with Google. To salvage the situation, we had to ask a friend who had been going to the same college as one of Google’s employees. He sent a few key emails on our behalf. Thankfully, within minutes, our ban was lifted. That’s only then that we learned why that happened. The Google bot had decided that the ad banner was too close to the game area.

The thing is, the ban lasted for 6 months. In that time, we didn’t pay ourselves just so we had enough money for our employee. That’s why we moved to Free to Play: to avoid another bad situation like that. That was in June 2012.

Here’s how we implemented it at first. In Transformice, you do not only collect cheese for the sake of it: you use it to buy hats for your mouse. We just gave the players the option to buy them with real money instead. And it’s been a huge success. On the first month only, we made €250,000.

Transformice relies on its massive multiplayer nature. We couldn’t add a barrier for people to get into the game. Because new players can get started within seconds, and for free, the community is really lively. Adding a subscription fee or making the game premium would have been a terrible idea. Transformice had been free for 2 years anyway. Changing that would have killed it. As hats and clothing is 100% optional, we were okay to let people pay for that.

How did you change the way you monetize over time?

The business model didn’t evolve for a long time. However, we just added watch to play: video ads on demand. We slowly adjusted things over time. For example, in 2014, we stopped giving away hats for free during events. As this was our main source of income, by giving them away, we hurt our sales. As a player, you only need so many hats for your mouse, and we gave a lot of them. This is one of the main changes we made, but aside from that, the business model didn’t change all that much.

The new video that reward the player with in-game currency only arrived in April, 2017.

Then, in a more general way, what are the biggest challenges you faced getting into Free to Play, and how did you overcome them?

I’m a bit ashamed to answer, because we didn’t really do our homework. To us, there were 2 ways to do Free to Play:

  • The bad way, like Zynga and other big actors, with energy-based systems that prevent you from playing if you don’t pay
  • And the right way, focused on cosmetics, as in League of Legends, with ethical prices

That was about all there was to know, for us, back at the time. We didn’t know what all these Key Performance Indicators were: ARPU, MAU, and all those barbaric acronyms. When we started doing Free to Play, we set our prices very low. Even today, our prices are much lower than other games on the market. The thing is, although you can lower your prices, you can’t increase them once they’re public. Players would not understand that.

It is hard to anticipate all those economic pitfalls, to understand all those metrics. They are not that accessible, not easy to grasp when you get started. We learned everything on the job, unfortunately. We should have learned all that in advance instead.

The in-game hat and skins shop

Over time, we added new paid options to cover different price points. Like, the ability to change the shaman’s crate, to buy complete outfits, or to rename your avatar. We started selling jewels that you can put on the tip of your mouse’s tail. They are shiny and they look unique, which justified a higher price than for hats and scarves.

The community accepts the fact that some services are more expensive than others. The idea that you will pay a higher fee to change your character’s name for example. We set this one around €10, which is standard in online games. It turns out it’s not a problem for people. We felt that was expensive, but it actually works.

I guess the community does that to support you at the same time?

Yes. At least when we first introduced Free to Play, our community wanted to support us financially. People were happy they could finally change their character’s fur color. They also sent us support messages, said they wanted to help after playing for 2 years for free. And we had donations before, but these didn’t work at all. It’s only when people could pay in exchange for something that they supported us massively. As I said before, in our first month, we made roughly €250 000.

Do use analytics for the game?

Yes, but not a lot. Laughs. Not as much as I would like to.

We use our own analysis code along with Grafana, a free program, to render the graphs. It’s a homemade solution that’s not very friendly.

As we have a multiplayer game, we always track the servers’ status. If we have issues, like many players dropping, or a long queue waiting to get into the server, we can act in response. Our metrics are technical, for the most part: we monitor the servers’ response time, or for example if we have too many payments that arrive at once. That’s one way to spot credit card fraud.

We use few metrics to improve our monetization. We look at the sales for the latest added products, to see if the players like them or not. But aside from that, it’s mainly global sales and the average basket. We didn’t design the system from the very start to track all the common KPIs. Now it’d be hard to do today. We would have to refactor the system heavily. We have 2 separate databases for payments and for the players, which are hard to link with one another. I do track retention however: how many accounts are created, and for how much time people stay around in the game. The percentage of players that stay for more than one hour, one day, one week… also what I call the “deserters”: players who come back after months of inactivity on the game.

Any advice for people looking to get into Free to Play games?

Do you homework! Laughs.

There are plenty of resources on the topic, and you want to read up so you don’t do the same mistakes others made before you. For example, for the KPIs, you don’t have to take everything. But you want to know what these are beforehand. You need to know how to use them, so you can pick the ones you need. You don’t want to use tons of metrics and never look at them. That’s worse than not having any at all. Whenever you add a metric your analytics board, think about what you’re going to do with it. Will it help you to take actions? You should think carefully, always have a plan before you start to track analytics.

Grafana is an open source graph generator. You send it streams of data using your analytics server, and use it to create your dashboard, from scratch.

How do you see Free to Play evolve in the coming years?

Free to Play already dominates over premium, if you look at it at a global scale. Be it in terms of revenue or the number of players, Free to Play is really big. So I don’t see it getting that much bigger. There will always be both premium and Free to Play games, as Free to Play is more adapted to certain genres. Especially multiplayer titles, when there’s a lot of interaction between the users.

How did creating your own company and Free to Play change your vision of game creation?

Back when I worked at Ankama, I had a narrow vision: I thought that everything would be better if I did it all myself and was in charge. Then I made my company and discovered it’s a lot more complicated than that. In particular, human resources. Clearly, the hardest part of game development is to manage people, to lead the entire team towards a common goal and not hurt anyone along the way. Keeping a great relationship with all our employees was the hardest thing I’ve had to learn to do.

That’s the real challenge. At least if you want to get it right. We’ve had clashes, and they were quite painful. If I could start again from the beginning, I would take management classes, and make sure to work on my leadership. Because at first, we didn’t have a common vision. We just let people do their work. But it takes many persons to ship a game. A lot of discussion, collaboration, and it is harder than you might think at first.

Dead Maze, Atelier 801’s next title, looks gorgeous

Atelier 801 is now working on their biggest project so far: Dead Maze, a post-apocalyptic, isometric MMO where players cooperate to survive to a zombie outbreak. You’ll get to collect resources, fight, craft to progress through its deadly, yet beautiful world. The first images from the game look gorgeous, and Melanie told us they’re looking to release it in open beta next summer. You can follow Dead Maze on Twitter.

Nathan Lovato

Nathan is a content crafter and Game Design & Analytics expert at GameAnalytics. When he's not doing research, you can find him writing and pulling colorful worlds out of his imagination.

Join a community of passionate game developers, who get our newsletter every week!

Sign up for a free surprise