YT YouTube2Text

YouTube2Text → Transcripts → The Cell Structure: How Supercell Turned the Traditional Org-Chart Upside Down

The Cell Structure: How Supercell Turned the Traditional Org-Chart Upside Down — Transcript

by GDC Festival of Gaming · 12,132 words · 1,792 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:05Hello. Uh good morning.
0:08Thank you all for coming. I I know it's
0:10it's quite early.
0:12So, my name is Ilkka. I'm from a company
0:14called Supercell and I will talk today
0:17about the way that we are organized. Um
0:21uh which we call the cell structure.
0:24But before I begin, I I want to
0:26emphasize that, you know, I'm not here
0:28today to tell you that or trying to
0:30convince you that this would somehow be
0:32the right culture or or the best
0:34culture.
0:35Uh obviously, it has worked out very
0:38well for us, but I believe I'm a firm
0:40believer that every company should have
0:42their own unique culture, which is like
0:45of course designed
0:46uh that company in mind.
0:49So, instead, my goal today and my sort
0:51of a hope is that that at least some of
0:53you guys would walk away from this room
0:55with maybe one or two new ideas how you
0:57can make
0:58your organization uh better and the
1:00culture culture better.
1:04So, I wanted to begin with a story about
1:07one of our our games.
1:09So, I think it was almost exactly um
1:136 years ago. I remember it quite well.
1:15We were in in this meeting room uh
1:19with most of the Supercellians at at the
1:21time and we were talking about new game
1:23ideas. So, at that time
1:26we had plans to release a couple of
1:27games that year, but we had no idea
1:30whether those games would be successful
1:32and and at that point like none of our
1:34games had been successful. So, we kind
1:36of we felt that we good to have like
1:38some game ideas in our back pocket.
1:41And
1:42one of the the ideas that was presented
1:44by this guy called Jon who presented his
1:47prototype which he called the Summoner
1:49and that that game essentially was a
1:51kind of real-time PvP game.
1:54And I remember people like really liking
1:56the prototype, but then like lots of
1:58concerns were expressed as well. And I
2:00remember myself for example saying that,
2:03you know, I just love what you've done
2:04here, but
2:05I just don't believe real-time PvP games
2:08would work on mobile. I wish it was
2:10asynchronous or something along those
2:11lines. And
2:13uh
2:13and and so this is like 6 years ago.
2:16And I remember that
2:17like uh
2:18when if you fast forward a couple of
2:20months from that moment, then the
2:21actually we we did get very lucky, but
2:23but our games Hay Day and Clash of Clans
2:25actually became relatively successful
2:27and and and we had all of our developers
2:29start with those two games. So, then we
2:31actually didn't have time to pursue any
2:32of the ideas that were presented in in
2:34that meeting.
2:36But then a couple of years went went by
2:39and and and and two other Supercell
2:41people Nikko and and Visa,
2:44they needed a game idea and they sort of
2:46remembered the Summoner prototype that
2:47was presented at that time. And they
2:49sort of picked it up and they were
2:51inspired about it and and they just
2:53created their own version of it and it
2:55was still was a real-time PvP game on on
2:59on on mobile. And again, there was sort
3:01of the same skepticism that was around
3:03the around the game.
3:06Uh but but despite that, they kind of
3:08pushed
3:09further and and and I even remember this
3:11moment with those two that we had a
3:13meeting and they came to me and said
3:14that they feel like like killing the
3:16game. And and I asked like so why why is
3:18that? And they said that well, she
3:20doesn't see that that many people
3:21believe in it. But then I asked that do
3:23you guys believe in it? And they said
3:24yes. And then I encouraged them to
3:26continue and and thankfully then they
3:28did. And but then like slowly but surely
3:31like
3:32like a group of like enthusiasts started
3:34to form around this game.
3:37But at some point the project um
3:39ran into some difficulties and and then
3:41Nikko decided that maybe like somebody
3:43else should lead this project. And then
3:45there was this guy Lasse
3:47who who was super super excited what had
3:49been done.
3:50But the only issue was that he believed
3:53that that the game should actually be
3:54like reworked completely and maybe they
3:57should go back to square one and and
3:58build it one more time. And and and on
4:01top of that, the team had the idea that
4:03instead of the IP that had been planned
4:05for this game
4:06code named Scroll,
4:07we should actually use the Clash of
4:09Clans IP. And you can just imagine like
4:11the amount of resistance from the people
4:14who had been originally involved with
4:16that game. And of course, Clash of Clans
4:18at that time was was our most valuable
4:20IP, so it was a big call. But you know,
4:23Lasse and the team didn't really care
4:24about the resistance. They just went on
4:26and and decided to do exactly that. And
4:29and what's interesting to me is that
4:31that even still a few months before the
4:33beta,
4:34there was quite a lot of people who
4:36thought that, you know, just doing this
4:37pure
4:39uh
4:39real-time PvP game, it's going to be
4:41like a really difficult thing to pull
4:43off on mobile. But but again, Lasse and
4:46and the team, they didn't care and and
4:47they just powered through and and
4:49released it on beta and and then
4:51that game is is is Clash Royale and and
4:54and obviously it was pretty pretty
4:55successful like from the launch.
4:58But then the story continues. So, fast
5:00forward again maybe 3 4 months from the
5:03launch and and we actually start to face
5:06some issues with the community. So, they
5:08were not happy with how the game worked.
5:10We had an issue like um
5:12uh what it was known as a trophy
5:14threshold. So, it had to do like with
5:16the progress of the players in the game.
5:18Uh players weren't happy about that for
5:20a good reason. It wasn't a good
5:21experience once you get it to a certain
5:23point in the game. And then what what
5:25really pissed our players off was this
5:27thing known as emotes in the in the
5:30game. So, what you can do in the game is
5:32that in your
5:33you can send emotes to your opponents.
5:35Um and and and lots of players were sort
5:37of using that to to basically irritate
5:39the the losing player and that really
5:41drove some of players like completely
5:43mad. And those complaints had been there
5:45from the very beginning, but the team
5:47the original team always thought that,
5:49you know, like this is part of the game.
5:51This is the soul of the game. It's a PvP
5:52game. So, we are not going to change it.
5:56But then
5:57Lasse wanted to do some other things and
5:58and and and and then a guy called
6:00Jonathan came
6:01and and and he he started to lead the
6:03team and there were some other changes.
6:05Uh and
6:07despite the fact that the all the
6:09original creators thought that this was
6:11this these emotes are an essential part
6:13of the game, they decided to make a call
6:16uh
6:17to enable the players to mute these
6:19emotes. And that of course like most
6:21many of the original team members like
6:23thought it was absolutely wrong. But
6:24yet, they did it. Uh and and because of
6:27that and and most of all because of some
6:29of the other changes that they had made
6:31to the progression of the game, the game
6:32started to grow quite quite quickly for
6:34the next 6 months and it is still at a
6:36very good place place today.
6:40So, um
6:41I think this story as a whole is a
6:43really great example how of our
6:44organization works about in good and
6:47bad. And and and it's a good example of
6:49our culture. And and but I don't think
6:51it's a good example of our culture. I'm
6:53not telling it because the game
6:55eventually was was successful. I'm
6:57telling it that, you know,
6:59this game game game together despite uh
7:03the fact that lots of people in the
7:04organization disagreed what was about to
7:06be done like starting from the CEO, uh
7:10many of the founders and many other
7:12uh sort of a people in the in the
7:14organization. But yet, somehow the game
7:16progressed and it was released and and
7:17and and was
7:19uh we were lucky in in a way that it it
7:20was successful.
7:22So, today I will talk about this this
7:25culture and and really like um trying to
7:27like tell you guys like what we have
7:28found out that what works well and then
7:30also like what what hasn't worked and
7:31what the challenges have been. And more
7:33than anything like what we have learned
7:35throughout these these years.
7:37But before I actually get into the
7:39actual culture and how it works, I want
7:40to tell a bit about like what actually
7:42led to the creation of it.
7:45So, so first of all,
7:47a bit about my background. So,
7:48unfortunately, I'm not the game
7:50developer. I I haven't developed any of
7:53our games and none of them have been my
7:55ideas. So, I can't take any credit for
7:57for the games.
7:58And even if I have always loved games
8:01and have been a big gamer,
8:02I actually ended up being part of the
8:04game industry just by an accident.
8:07And it's it's funny to think this, but
8:08originally actually I for whatever
8:10reason I wanted to be an when I went to
8:12the university, I wanted to be an
8:13investment banker or management
8:15consultant. Until today, I know don't
8:17know why. I'm sure I'm very happy it
8:19didn't happen eventually.
8:21Um and the reason why it didn't happen
8:23is that I got really excited about the
8:24entrepreneurship at some point and and
8:27and I was lucky enough that I bumped
8:29into this group of people and they
8:30wanted to found their own game company.
8:33Uh and it and it just happened to be a
8:35game company. It could have been any
8:36other company like from my point of
8:37view. But these guys wanted to do games
8:39and they were all game developers.
8:42But none of them were like interested in
8:44in in doing any kind of
8:45or handling any of the commercial or the
8:47business side of the of of running a
8:50business. And and they they couldn't
8:52afford to you know pay any salaries and
8:54and I'm sure they didn't have any other
8:56applicants than myself for the job. So,
8:58then I finally got the job and and and
9:00and became the CEO of the company. And
9:02and that company was called Sumea and
9:03this is like year 2000. And funnily
9:05enough, we were actually doing or our
9:07plan was to do mobile games already at
9:10at that time.
9:12Uh
9:13eventually, the company did okay and it
9:15actually was sold to a company called
9:16Digital Chocolate, which is actually
9:18from here based was from here based in
9:20San Mateo. And then my job was to run
9:23the European operations of the of the
9:25company.
9:26Uh and you know, those times were quite
9:29different. So, the ones of you who are
9:30used to or all of us who are used to
9:32like dealing with the app stores, I mean
9:34this was a very different time because
9:36the distribution was through like
9:37carriers and and you had to do like
9:39these separate distribution deals with
9:41every single one of them. And
9:42and it was quite interesting. I actually
9:45found this
9:46photo from from those times. As you can
9:48see, it's a very serious business
9:51with a jacket and everything. And just
9:52look at that phone like on my lap like
9:55uh but the funniest thing about those
9:57days is that you were basically selling
9:58games to people who didn't really care
10:00about games. Most of them they didn't
10:02they didn't play the games and it was
10:04just, you know, it's just traditional
10:05sales work. I I can't say that I would
10:07miss miss those times uh days too too
10:09much.
10:10But then uh after that I my job was to
10:12actually run the global studios
10:13organization
10:15of digital chocolate and and we had
10:17three studios and I guess around 200
10:19people all together in those those
10:20studios.
10:22And uh
10:24So I come from my told you guys my I
10:26come from this quite analytical
10:27background. Um, like and and sort of
10:29myself said my my for some reason my my
10:31dream job was to be as a something like
10:33management consultant. Uh,
10:36and and so I
10:37originally took a very kind of what I
10:39thought was a very analytical and
10:40thoughtful
10:42uh approach to design the organization.
10:44And by the way like uh since I've sort
10:46of created this organization myself,
10:49uh I I I believe that I also have a
10:50right to kind of criticize it and and
10:53you know all all of them funny things
10:55and mistakes that I'm going to talk
10:56about I mean I
10:58those are all like my my mistakes and
11:01uh and the other thing is that that even
11:03if if some of this stuff will will sound
11:05quite funny, I mean keep in mind that
11:07everything was done like with the best
11:08intentions. So already at that time uh
11:11our goal was to just simply to create
11:13the the best best games.
11:16But anyway so we my my approach was that
11:19I wanted to the company to be very well
11:21organized. So like every everybody
11:23belonged to their kind of own unit or
11:25box and and everybody reported to some
11:27somebody and they had this like very
11:29well thought out matrix organization.
11:32And and they had process for every
11:33single thing you can imagine like from
11:35creating new games to to many many other
11:38things and and everything sort of looked
11:40really great on on on paper.
11:42And and most importantly
11:44uh
11:45what we liked about their indicators is
11:46that who has the product vision like who
11:48ultimately decides what type of games
11:51the company does. And and the product
11:52vision was very much at the top. So it
11:54was myself, it were the three head of
11:56the studios and some product managers
11:59and and our small group essentially
12:00controlled what the company was was
12:02doing.
12:04Uh and then like uh
12:06what made it probably even worse is that
12:07then the organization grew so we hired
12:10all of these people and we opened these
12:11new studios. Our approach to manage that
12:14growth was actually to like even in
12:17further increase the control so we added
12:19layers of management and we added the
12:22amount of uh added new processes to the
12:24company because we wanted to control it.
12:27And uh
12:29I've I thought a lot like in the recent
12:32years like uh what sort of drove uh
12:36all of those things and and you know it
12:37really was like this feeling that
12:39I I guess ultimately I had to have that
12:41you know I wanted to feel like that I am
12:43in control of our destiny and you know
12:45having everything all of these written
12:47plans and and and and and org charts and
12:50everything and processes it kind of made
12:51me feel like that we would somehow be in
12:54control of our our destiny. That's
12:56probably the best explanation that I've
12:58come up with.
12:59And we exercise this control through
13:02through a couple of processes.
13:04I'm going to give you example of two.
13:07The first example is is is which I think
13:10is relatively widely used in in in the
13:12games industry one form or another is
13:14this kind of green light process. We
13:16call it the YDC meaning that why should
13:19digital chocolate build this product.
13:21Essentially it's a mini business plan
13:23for every game. Uh
13:25so basically starting with things like
13:27you know what's the market opportunity
13:29like uh what's the size of opportunity
13:31what does the competition look like how
13:34are our games different and or better
13:37you know what's the cost budget what are
13:39risks so forth so forth. And and and
13:41from every single new game this type of
13:43document product 10 pages um
13:46was was required.
13:48And uh and to make it worse I had this
13:50idea which I thought was a good idea but
13:52it resulted to be a very bad idea. Uh
13:55but
13:56so my goal was that I wanted to like
13:57make everybody in the company feel all
14:00the departments feel that that they are
14:02sort of a part of the process and and
14:04they can and sort of create a sense of
14:06ownership. So
14:08I required like the the game developers
14:10actually talk to heads of each and every
14:12department and sort of to get their sign
14:14off or support for this new game idea.
14:16So these people would talk to the head
14:18of the finance, head of the marketing,
14:20head of sales, head of operations to
14:22kind of get their approval. And and you
14:24can sort of guess where it leads to.
14:26Like uh in in these meetings that they
14:28had about the green light uh decisions
14:30like there was no real discussion
14:31because the real discussion had been
14:33like before the meeting and and it was
14:35like super political process where like
14:38the developers would do anything they
14:39could possibly can to get an approval
14:41from certain people.
14:43Uh and and then like and if if if there
14:46was no approval from a person then that
14:48game wouldn't the idea wouldn't even
14:49appear in in this meeting. So the value
14:51of the meeting was was close to close to
14:53zero.
14:55Uh
14:56And also like some other problems that
14:57that it had was that you know like if
14:59you think about it like I mean I mean
15:01nobody has a crystal ball. It's really
15:03hard to predict which games will be
15:04successful. So then you could make the
15:06argument that you know why would this
15:08group of people have that crystal ball
15:10and why should they be making the
15:11decisions. Uh and and one could argue
15:14that who really knows the games and and
15:16the players best it's probably is the
15:17game developers themselves.
15:20Also like this process it really favored
15:22sort of what I would call like
15:23incremental innovation. Like very
15:27kind of products that are little better
15:28than the existing ones. But and and and
15:30the reason is that for you to make a
15:32case like what type of product to build
15:34the easiest way to build your financial
15:37estimate was to look at what already
15:38exists in the market and say that okay
15:40we are better because of these games
15:42because X Y and Z and therefore the
15:45revenue estimate it's like we're going
15:47to capture X percentage share of this
15:49market or something like that. And that
15:51was sort of an
15:52easiest way to sell this stuff to people
15:54who didn't come from games background.
15:55So for example a guy with a financial
15:57background. I mean that's the language
15:59that that person understands.
16:02But what
16:03what the problem is that this type of
16:04process doesn't favor the truly innova-
16:07innovative ideas. Like if you really
16:08want to build something from scratch you
16:10want to build something that doesn't
16:12exist yet you can't build a case like
16:14that around it. And and surprise
16:15surprise we really didn't have many of
16:17those type of games like coming through
16:19the the process.
16:21And then the the the um
16:23the last uh sort of a problem with the
16:25uh process was that just you know it
16:27required a lots of time from the
16:28developers. So instead of them like
16:30writing prototypes and and and you know
16:32building the actual games they were
16:34building pitches which of course isn't a
16:36very good use of time. They could use
16:38that time for the products themselves
16:40instead.
16:42So that would be one example of a
16:43process. Um
16:45Another example is is like that we at
16:47some point we introduced this like
16:48monthly review uh processes and and we
16:52uh they were essentially like status
16:54reports from the game teams to the upper
16:56management like what's going on. And and
16:58the purpose was like that we wanted to
17:00feel in control that and make sure that
17:02these games are actually progressing as
17:04as we had agreed and and you know are
17:06still shooting towards the vision that
17:07we had had agreed. And and and what
17:10really like made these meetings bad was
17:12that they they were also essentially
17:14like go no go decision points. So they
17:16also were like green light meetings as
17:17well meaning that that it it could be
17:20that your game gets discontinued uh in
17:23in in this meeting.
17:24And of course you can imagine what
17:25happens. Of course the game teams would
17:27come to this meeting very well prepared.
17:29They had done all of this work the prior
17:31week to prepare for the meeting. They
17:33had created a lot of materials just for
17:35purposes of this meeting. And after the
17:37meeting uh
17:39had the game continued they would just
17:40throw away all of that stuff and
17:42continue where they left off. And at
17:43some point I think I figured out that
17:45like some teams they literally used one
17:47week of every single month just for the
17:49purposes of of this meeting. Of course
17:51of course that doesn't make any any
17:53sense.
17:53Also you can imagine the amount of
17:55frustration that these developers had
17:57when they had to explain uh game stuff
18:00and game development stuff to people who
18:02like weren't like game developers at
18:04all. Like again like uh you know it was
18:07my fault. I was the one who invited this
18:09like finance and marketing and sales
18:11guys to the to the meeting.
18:13Uh it's definitely not their fault but
18:14you know of course they had questions
18:16and you know it wasn't very productive.
18:18And and the last thing about this
18:20meeting that again you can imagine the
18:22large group of people and very easily
18:24lead to this kind of sign by committee
18:26type of thinking. So the ideas that
18:27would go through are the ideas that
18:29doesn't get anybody irritated but then
18:32the other hand you know aren't really
18:34that great or innovative either because
18:36like most of the the big innovative
18:37ideas they are
18:39uh things like for example Clash Royale
18:41that did something new but then you
18:42don't really understand it before it's
18:44it's out there.
18:47And and there were other uh and I could
18:49have like a
18:50pick some other processes as well but I
18:52these are probably the big top two ones
18:53on on on my list which really didn't
18:55work out.
18:56And and the really really interesting
18:58thing about all of this is and all of
19:00these processes I said they were put
19:02together with best intentions. And but
19:03the thing is that despite all of this I
19:06mean believe it or not but we were
19:07actually able to put out some really
19:09great games. So ones of you who who were
19:12part of the mobile games industry back
19:14back in those days might remember a game
19:16called Tower Blocks which was a really
19:18big hit for us later on when Facebook
19:20just started their gaming platform we we
19:23uh released a game called Millionaire
19:24City that that got a huge amount of
19:26users and there were others as as well.
19:29And at some point I started to think
19:31that what really is the common nominator
19:33uh behind all of these games.
19:36And I the more I thought about it more I
19:38realized that actually like
19:40first of all like all of these games
19:42they had a really great team behind uh
19:45them.
19:45Then two they were released exactly at
19:48the right time. So we got the timing
19:50right which of course is as mostly to do
19:54with with luck.
19:55But then thirdly, it also seemed to me
19:57that what explained the success is that
20:00the less the management, like myself and
20:02some others, less the management had to
20:04do with the game, the better were the
20:07chances of success. And and somehow
20:09these games, these hit games, they just
20:11happened, like almost like accidents.
20:14And and I once actually joked to
20:15somebody that that that we were able to
20:18like put out these hit games despite all
20:21of these processes that we had in place.
20:23And that of course like tells you
20:25something like how great these teams and
20:27the people there at at the company.
20:30And then I like sort of finally got it
20:31and it's it's really embarrassing to
20:33almost to talk about it because it took
20:34so so many years, but it I finally got
20:36that these processes, like no matter how
20:39well thought they thought out they are,
20:41they just do not translate into great
20:44games.
20:45And then I the kind of thought process
20:47continued and and started to think that
20:49if the success truly is and is only
20:51about the team and the timing and the
20:54timing you can't really control because
20:56that's about luck, but what you can
20:58control is actually the team.
21:00And if it really is like about the
21:03getting the best possible team together
21:06and and you know having an environment
21:08that really like doesn't slow these
21:09teams down, so there's like none of this
21:11management stuff
21:13like
21:14it creates friction for the team. If
21:16it's all about the team, like what if
21:18what would happen if you would sort of
21:20build a kind of new type of company, but
21:21would only be about people and and and
21:23these teams and and having an
21:25environment where there is no friction.
21:27And and what if actually like a
21:29what what if the
21:31the the people who would be sort of a
21:32superstars of the organization would be
21:34the game teams themselves and not sort
21:36of the the founders or or the the
21:38management. And what would happen if it
21:40was actually the teams who owned the the
21:43product vision?
21:44And then at some point we had this idea
21:46that what would happen if we actually
21:48turn this organization chart upside
21:50down. And and and actually like it would
21:52be the game teams themselves who kind of
21:54owned their game. And that's what the
21:56light bulb up there indicates.
22:00And and then we kind of took this idea
22:01even further and thought that uh
22:03we could actually think these teams as
22:06as they're kind of almost like their own
22:07independent companies inside the the
22:10greater company. And then at some point
22:13somebody suggested that since this is
22:14such a radical idea, I mean let's not
22:16call them teams, let's call them cells.
22:19And then somebody asked that okay, like
22:21if that's the case, like what is this
22:23whole thing called? And then somebody
22:24came up with a word Supercell and that's
22:26where the kind of name of the company
22:27comes from.
22:29But really the the kind of the the the
22:31idea here is that there is no central
22:33process. I mean every team can be
22:36different. And and and you know like
22:38that enables us to hire from much
22:40broader base of people because you don't
22:42have to be like this just like one one
22:45type of
22:46person to be to get to get hired. And
22:48and every team can have their own ways
22:50of working.
22:51And and we sort of
22:52wanted to see that as a source of
22:54strength rather than a source of of
22:57weakness.
22:59Well, then at at some some point
23:00somebody asked that okay, but
23:02if it's all about the teams, like what
23:04do you do then? Like what is the role of
23:06of management in in this type of case?
23:08And and we thought about it a lot and
23:10and and and really the the the answer
23:13seems to be that it's it's our job to to
23:15put together the best possible teams and
23:18it's our job to make sure that these
23:21teams have the best possible environment
23:22to work in.
23:24And by best possible environment I don't
23:25mean any I don't mean a fancy office,
23:28for example. I mean I mean an
23:29environment where these teams can have
23:31the biggest possible impact and there's
23:33no friction in in the organization.
23:37Also one of the things that we kind of
23:38got excited about was this idea that if
23:40these teams make the decisions, it
23:42probably like will increase the the kind
23:44of speed of execution, but also because
23:47they don't need approvals, but it will
23:49also increase the quality
23:51of the decisions because then the
23:54decisions get to be made as close to the
23:56player as possible. Because it's
23:58obviously these guys who know the games
23:59best and they probably also know their
24:01users or the players the best, so it
24:03just makes makes sense.
24:05Uh
24:06and and and these days I I I think that
24:08our organization it it works the best uh
24:12then like I make as few decisions as
24:14possible. Like and you know I have
24:16almost
24:17we've sort of made it a goal that that
24:19most of the decisions, if not all of
24:21them, they should be made at the team
24:22level, not at that sort of the
24:24management level.
24:25And and you know to describe that goal
24:27I've even like
24:28started to say to our people that my
24:30goal is to be actually the least
24:31powerful CEO. And I'm meaning that you
24:33know the less decisions I make, the
24:35better the organization works.
24:37And sometimes this sort of leads to some
24:39really interesting situations. So
24:42uh
24:42for example, when we we work with some
24:45uh
24:45uh outside outside companies, like for
24:47example with partners uh
24:50uh platform partners, for example, like
24:52it it it always takes them a while to
24:54realize that at Supercell the most
24:56important meetings you can possibly
24:58have, they are with the game teams. So
25:00for example, if they're meeting with me,
25:02it in most cases it's waste of their
25:04time because at the end of the day it's
25:06not me who's making these decisions. And
25:08they usually don't believe it the first
25:10or even the second time we tell that to
25:12them, but maybe after third or fourth
25:14time and then they start to see like how
25:15we work,
25:17it's suddenly they they they sort of
25:19they get it.
25:20Or or there are like things like uh
25:23recently we when it was the
25:25International Day of the Girl, we part
25:28or I participated in the in this
25:30initiative called Girls Takeover. So the
25:32idea is that this uh say a teenage age
25:35girl comes and you know takes over the
25:37job of the CEO and and makes all the
25:40decisions for for one day. So
25:42and they asked me to participate and
25:44because I thought this is a very
25:45important initiative, of course I wanted
25:47to participate, but I kind of pre-warned
25:50them all all of these girls, Sonia, that
25:52that I I'm happy to support you and we
25:55will do this, but I just want to warn
25:57you that you might be actually quite
25:58disappointed because the fact is that I
26:00make very few decisions in this company.
26:03But you know she took the chance anyway
26:06and and and I think she did the right
26:08thing by touring the studio and and
26:10talking to the game teams, which is like
26:11the always the most interesting part of
26:13of my day. But you know
26:15it is just interesting like how how this
26:17type of way of organizing it it changes
26:19how you also deal with some of the
26:20external parties.
26:24One one other idea
26:26that that's that's very close to our
26:27culture is this idea of of the power of
26:30a small team. So at the previous company
26:33we we had like seen in our our own eyes
26:35that how powerful these like these very
26:38small group of people can be if they are
26:40just the right people and and work well
26:42together. It almost seemed that uh
26:45that
26:46uh and it's counterintuitive, but the
26:48scarcity of resources, meaning the
26:50number of people it it sort of uh
26:53it it leads to
26:55better prioritization, better focus and
26:58then that in turn leads to better
26:59quality. And and you know we we see saw
27:01it at the previous company time after
27:03time and also we've seen it at
27:04Supercell. So quite often when things
27:06are aren't working out for us, one of
27:08the de facto solutions that we have is
27:10that let's make the team a bit smaller.
27:12Uh and and and and you know let's try to
27:14put together the smallest possible team.
27:20So um it probably isn't a surprise that
27:22that in our at Supercell we don't have
27:24any any kind of green light process. But
27:25then the question is that that gets
27:27often asked that how do you guys decide
27:29what games to build?
27:32How do you sort of quote control it? And
27:35and actually like the only thing that we
27:36sort of control
27:38is the team and the team's composition
27:40and and you know of course I have to be
27:42able to trust the lead of the team. But
27:44after that, you know it's up to the team
27:46to develop the games they want as long
27:47as they develop the games in a framework
27:50we've agreed on. I mean our dream is to
27:51create these games that people would
27:53play for years and that are targeted for
27:55a widest possible audience. But that's
27:58really is the framework and within that
27:59framework the team is is free to do what
28:01they want and build the game that they
28:03want.
28:04Also
28:06we are known
28:07for like killing a lot of our games. So
28:09how do we how do we do that then?
28:12Well, at Supercell I would say that two
28:14entities have like
28:16control. So one is the game development
28:18team itself and then the other one is
28:20players. So and the team is in control
28:23in full control during the the
28:25development phase. So when we still
28:27don't have any
28:28hard data or metrics, it's it's the team
28:31and only the team who can decide
28:33whether the game will continue or not
28:35continue and be killed.
28:37But then like when the game goes from
28:40the development phase to a soft launch
28:41phase, so we or the beta phase, and and
28:44we start to actually get some like uh
28:46metrics on the players, then the control
28:49shifts in a way from uh
28:52uh from the team to the players because
28:54then it really is up to the players. And
28:55of course we will be observing the
28:57metrics and if they're good enough, we
28:59will uh
29:00uh launch the game globally. And what's
29:02good enough? Well, that's something that
29:03we always decide before the game goes to
29:06beta and we tell the metrics to
29:07everybody in the company just to make
29:09sure that we are honest to ourselves.
29:11And after that it's actually quite
29:12simple. Like if the game reaches the
29:14metric goals, we will release it to
29:15global and if it doesn't, we'll kill it
29:17and we'll we'll do something else.
29:21Uh we
29:23this this model has a couple of couple
29:24of like challenges that are important to
29:26understand. Um
29:28The first challenge is it's an
29:29interesting one. I mean the thing is
29:30that lots of people do not believe that
29:32this is true. Like not even after
29:34they've joined, at least for the first
29:36month or so. Uh
29:38many people outside the company believe
29:40that this is some kind of a PR trick. So
29:42you know people like me would come to
29:44conferences like this in the hope to
29:46talk about it and then it's attract
29:48talent to the to the company, but
29:50and and it's just a PR thing and it
29:52doesn't really isn't true. Um but
29:54but that's obviously not not the case
29:57and it wouldn't make any sense
29:59uh because people discover it very
30:00quickly when they when they would would
30:02join us. And it has happened to me like
30:04so often that you know, somebody comes
30:06in
30:06uh
30:07is with the roster for a couple of weeks
30:09and I meet that somebody at the coffee
30:10machine and and and that person goes,
30:12"Hey Ilkka, I have to confess you
30:14something that I actually didn't believe
30:16like this independence thing
30:18when when I joined, but I but now I've
30:20realized that it actually is is true."
30:23The other thing uh that is challenge for
30:25us that like outside looking in there's
30:27this misconception about the the the
30:29model. There are lots of people think
30:31that a Supercell is is somehow this just
30:34this small happy family where anybody
30:36can do whatever they want and not care
30:38about the the results. And that really
30:40isn't true at all actually. Um I mean,
30:43what actually keeps this whole thing
30:46together and act as a glue is this our
30:48very high quality bar. So, as I just
30:50said, you know, our goal is to create
30:52these games that people would play for
30:54years and years and years. And that's a
30:55very high bar as as you guys know. And
30:58and then if you combine that with a
30:59completely transparent information
31:02sharing, so every single team knows
31:04exactly how all the other teams are
31:06doing
31:07just because we share all the metrics to
31:09everybody in the company. And also these
31:11teams have to give updates to the rest
31:13of the company quite regularly. So, uh
31:15so and and it's very apparent like how
31:17the other teams are doing. So, so the
31:19combination of the high quality bar and
31:21that is very transparent
31:23makes it sort of a high performance
31:24organization where people really want to
31:26like always
31:27uh get to the bar and and and and even
31:30even higher.
31:31And and you know, as I said, like
31:32whatever doesn't meet the bar, it
31:34actually gets killed, which of course is
31:35always a hard hard decision.
31:38And and you know, like in our
31:39organization the fun part to talk about
31:41this always the independence thing. I
31:42mean, of course everybody wants to be a
31:44part of an independent team or small
31:46team that can do what they want. The
31:47harder part is to talk about the other
31:50side of the same coin, which is
31:52responsibility. And and the way we think
31:54about it that every single member of a
31:56team, they are responsible towards each
31:58other and to all the other
32:00Supercellians. And that's a that's a
32:01very very big responsibility. And and
32:04and it also means that in our in the our
32:07culture, we don't think that the
32:09the responsibility lies with the
32:11management. I mean, if there's a
32:13problem, we expect the teams themselves
32:15to do something about it and not become
32:17of a management to to to interfere.
32:21Uh
32:22And and and you know, that actually
32:23results in in in a lot of like killing
32:25decisions. So, like I mean, I I believe
32:27it's actually easier for people to kill
32:30their work if it's not working out then
32:32they can when they are in full control
32:33of that
32:34by by themselves and the decision is is
32:37is is theirs. So, um
32:39uh for example, I a guy called Jonathan
32:42um he I think he was here speaking like
32:442 years ago about a game called Smash
32:46Land, which he was the lead of and game
32:48that we dearly loved and and uh uh
32:51and and he was telling the story here in
32:53the same conference about how he they
32:56they uh typical Finnish way, even if
32:58Jonathan comes from New Zealand, I think
33:00they had a meeting in a sauna with the
33:02team and then they had uh and the game
33:04was in beta and it was almost reaching
33:06the beta metrics, but just almost, not
33:08quite.
33:09Uh and and they had had this show of
33:11hands that okay, like how many of the
33:13the people in the sauna believe that
33:15this is the best game they will ever do.
33:17And I think there was very little hands.
33:19Uh so, which in which and then they
33:21decided in that evening to kill the
33:23game. And then I think Jonathan sent
33:25this email to everybody in the company
33:27informing. And I remember that very well
33:28because I happened to be traveling here
33:30and I was checking my emails in the
33:32morning. I said, "Okay, like they they
33:33actually decided this game to kill this
33:36game." And I I felt really sad about it
33:38because one, I loved the game myself. I
33:40I played it a lot and whenever I went to
33:42home, I I I played it a lot with my
33:45kids. And I said, "Okay, how I'm going
33:46to how I'm going to explain it to this
33:47to my kids?"
33:48Uh two,
33:50I felt really sad for the guys and for
33:52the team. I know they had worked super
33:54super hard for for that game and and
33:57then they had to kill it. But then I
33:58also felt very proud. Like I mean, just
34:00think about like it's amazing that that
34:02these people have like poured their
34:04heart and soul to this game that
34:05everybody in the company loves and it's
34:07very close to these metrics
34:09and yet they think that they can do
34:11better and they kill it. So, I just felt
34:12proud about the uh the the those guys uh
34:16and and the team at at that time. And
34:18you know, we do kill a lot of games and
34:19we are known for that. So, between our
34:21game number three and four, so between
34:23Boom Beach and Clash Royale, I believe
34:25we had developed 10 games to relatively
34:28advanced state and we out of those 10 we
34:30killed nine. And the one that
34:32uh we didn't kill was was Clash Royale.
34:34And and of course like killing games,
34:36it's definitely not fun and failing is
34:38not fun. Uh
34:40but still like we
34:41we've actually tried to take some
34:42inspiration from Pixar. They they often
34:45say that failure is an essential and
34:48inevitable part of the creative process.
34:50So, I mean like without failure, there
34:52wouldn't be any hit games either. And
34:54and you know, if you're not failing, it
34:56just means that you're not your bar
34:57isn't high enough and you're not taking
34:59enough risk. And that's how we try to
35:01think about it.
35:02It doesn't make failing any any easier,
35:04but but you know, that's the kind of a
35:06the the
35:07our way of thought about it.
35:10Okay, um
35:11um
35:13I I want to like uh close my talking
35:15about some of the learnings uh
35:17uh that we had over the years and some
35:19of the challenges uh
35:22So, and this will sound self-evident,
35:24but I still need to say it. Uh I mean,
35:26this model is exactly as good or as bad
35:31as the teams are. It it really is all
35:33about the teams.
35:35And and like from this perspective, it's
35:37actually it's interesting. Like we
35:38actually like originally when they
35:40founded the company, we didn't say that
35:42the best teams make the best games. We
35:44said that the the best people make the
35:46best games. And it took us like many
35:48years to actually realize that actually
35:50it's not about the people, it's it
35:52really is about the teams. And and you
35:54know, putting together like a great
35:56teams is super hard. And and and to sort
35:58of use a sports analogy, like what we
36:00realized over the years is that the best
36:02team, it's not necessarily the one with
36:04the best individual players,
36:07but the team that plays best together.
36:09And and you know, putting together such
36:11a team, like it's it's it's not easy and
36:14I I don't know if there's even lots of
36:16people have tried to put some science
36:17around it, but but still at least in our
36:19experience, the only way to really know
36:21how the team is working, it's just to
36:23put together the team and see how how
36:24they work together.
36:27Uh the other challenge is that you know,
36:29I talked a lot about how we love the the
36:31kind of small size of the teams.
36:33And and and yes, there's a lot of great
36:35things about it. When you have a small
36:36team, you don't really need like any
36:38layers of hierarchy in the team. You
36:39don't need process, you know, everything
36:41can be planned on a whiteboard. You need
36:43very little meetings because everybody
36:45can hear each other out. They're working
36:47in the same room. You know, the the size
36:49of our new games teams are like two to
36:51seven people. Live games maybe 10 to 17
36:54people um to give you guys an idea like
36:56how small these teams are. And as I
36:58said, we believe that the all of this
37:00actually leads to better quality. And
37:01it's a way more fun to work in in in a
37:04smaller team.
37:05Uh but it also creates challenges.
37:08One really interesting challenge that we
37:09actually didn't realize early on is that
37:11it actually like for some people it it
37:13can create a relatively stressful
37:15environment. So, the combination of like
37:17having a very high quality bar and then
37:20this idea that there's like no it's only
37:22about you because there are no
37:24processes. It really is about only about
37:26you and the team, of course.
37:28Uh and then that combined with the small
37:29team size, so you know, everybody
37:31basically is designing a game and you
37:32know, there are very little like
37:33specialization like
37:36uh in in the team. So, so that
37:37combination like that can become
37:39stressful. And you know, we've actually
37:41done a quite a lot of work to like make
37:43that stress go away, but that's a hard
37:44thing to do.
37:46And then the other um
37:48thing that that is uh
37:50a challenge for us and I don't think
37:51we've like have been very good at
37:53solving this so far is that how how do
37:55we kind of create content fast enough
37:56for our players given the small size of
37:59the team?
38:02Another a double-edged sword is is is
38:05you know, this idea that you have very
38:06little like visible structure. And
38:08again, there's like tons of things that
38:09we love about it, but the challenge that
38:11comes is that the
38:12uh
38:13it sort of may it may seem
38:16uh chaotic uh like especially early on
38:18and and you know, we
38:20had had have had a lot quite a learned
38:22quite a lot of lessons for example about
38:24onboarding. Like like
38:26many years back, say 4 years back, it
38:28was probably like a really difficult to
38:30join Supercell because you would just
38:31come in and somebody might in the best
38:33case show where you there's your seat
38:35and give you the laptop. And but then,
38:36you know, when you're on your own. And
38:38and and but then finally they realized
38:40that they really have to invest to this
38:41and now we are like way way way better
38:44at at that. But you know, like
38:47in certain uh
38:48places you have to have structure and it
38:50and it's especially important for the
38:52for the new people.
38:54Uh sometimes people ask I mean, we still
38:56don't have any HR. So, we do have
38:58recruiting and of course we have office
39:00people and admin people, but we don't
39:02have like HR people.
39:04Uh and and people ask that if people and
39:06the teams are so important to Supercell,
39:08like why on earth don't you have HR
39:11uh person um or somebody who's
39:13responsible for HR? And my response
39:16always is that well, that is exactly
39:17because people are so important to us.
39:19So, I don't believe that you can somehow
39:21outsource HR to a third person. I mean,
39:24we if there are any kind of HR issues,
39:26we want them to be dealt with between
39:27the lead and that person themselves. I
39:29don't we don't want to have like a third
39:31people
39:32um uh
39:34in involved there.
39:35But it creates some challenges at at at
39:37times.
39:40Also, as I'm sure you've already figured
39:42out, like it's almost impossible to
39:44reach consensus in a culture like like
39:47ours.
39:48And and you know, like many of our
39:50games, I would say most of our games, we
39:52have like four four games have faced
39:54like strong
39:56kind of
39:57I would call it almost like internal
39:58opposition or people who have just like
40:00disagreed
40:02about the direction of the game and and
40:04you know, have fought fought that we
40:05shouldn't even be building this game in
40:07the first place. And
40:08just talk about Clash Royale and all the
40:11twists and turns. The other example
40:13could be a game called Boom Beach, which
40:14was our game number three. I still
40:16remember this meeting there, myself and
40:19all the game leads. So, every leader of
40:21a game team sat, including the lead of a
40:23Boom Beach. And and and we had a like a
40:27like very heated discussion about Boom
40:29Beach and people were not happy about it
40:30at the time. And then we like again had
40:33a show of hands that okay, how many
40:34people believe that this game should be
40:36killed? And every single person in that
40:39room, with the exception of the Boom
40:41Beach lead, wanted to kill that game.
40:44And then we we had this really
40:45interesting discussion that okay, like
40:47like all of us except you, like all of
40:50us we want to kill this game and we
40:51don't believe in it. But then again,
40:53like if we now kill it, like at the same
40:56time, not only do we kill the game, but
40:57we will also kill the culture. Because I
40:59mean, after that, you know, we couldn't
41:01say that Supercell is all about these
41:03small and independent teams who like are
41:05responsible.
41:06And if we make the this decision to kill
41:09this game, all of a sudden the
41:10responsibility is not with the team
41:12anymore, it shifts to us. Because then
41:14it would be us who who who make the make
41:17the call. And then, you know, we we
41:19decided that okay, out of these two
41:20options, like we are still we are here
41:24in for the long term and culture is so
41:25important. So, you know, you guys should
41:27just do the game. And then again,
41:29thankfully they did and it's it's it's
41:30been a big big hit for us.
41:33So, we start try to like
41:37and we borrowed this from Jeff Bezos
41:39from Amazon this phrase that disagree
41:41and commit. But you know, like we try to
41:43like enforce this idea that even if you
41:45disagree, you can still sort of stand
41:48behind the decision. And and you know,
41:50and and the reason why you can stand
41:52behind the decision is that you trust
41:55the team who has made that decision,
41:56even if you disagree.
41:58So, so that's that's how we can think
42:00about it these days. But still we have
42:02these heated discussions and and I think
42:04they're really good and and essential
42:05part of the the culture.
42:09The last learning is is this one.
42:12Uh
42:13It it really is interesting, but this
42:15traditional approach and which and by
42:17traditional approach I
42:20I'm talking about stuff like top desire
42:22to do like top-down management, desire
42:24to control stuff like from outside the
42:26team,
42:27desire to create rules and processes and
42:30and you know, all the stuff that I just
42:31talked about. It's it's almost like
42:33gravity. Like you know, it it sort of it
42:36pulls you like all the time like towards
42:38it and and you have to like almost fight
42:41back like every single day that it
42:43doesn't happen. And it happens in in in
42:45in all kinds of
42:47situations. And
42:49and and and the interesting thing is
42:51that
42:52that like fighting this it gets harder
42:56and harder the more successful you are.
42:59And and like for example, in our case,
43:01like the more successful we become,
43:03because we've obviously been like
43:04ridiculously lucky like in in with our
43:07four games.
43:08But the more successful these games have
43:10become, like there's always like there's
43:12some kind of pressure and natural kind
43:14of instinct to do even more stuff and
43:16grow the team sizes. And there all kinds
43:18of opportunities that you say no to need
43:20to say no to. And and and we have to
43:22really like like push ourselves to be be
43:25always like saying no to all kinds of
43:26things so that we could stay laser
43:28focused on the on the games.
43:31Also, what's even even harder is that
43:33you know, it's easy to talk about this
43:35stuff when things are going well. But
43:37when things aren't going well, it's it's
43:40it's way harder. Because there's this
43:42like thing that comes to you that that
43:44this kind of pressure to kind of get the
43:46control back from the teams and give it
43:48to some other people. There will be
43:50people who will say that okay, we'll
43:51we'll come and we'll fix it for this for
43:53you. And and and you know, lots of like
43:55opinionated people as as I guess most of
43:57the great game developers are. So,
43:59there's really strong opinions.
44:01And and it's it's it's and we really
44:03have to fight
44:05strongly against this this idea that for
44:07example, there are like some some really
44:09really great people in the organization
44:11who've done a great great work hit games
44:14before. And if they feel strongly about
44:16the game that they are not part of, like
44:18who gets to decide? And we've always
44:20taken the kind of side of the team. And
44:22and and we've tried to resist the
44:24temptation to give like control to some
44:27people outside the team. But it's a
44:28discussion that that we have like
44:30surprisingly often. And and and as I
44:32said, you know, it gets even harder when
44:34things aren't going that well.
44:36Also, like you you really have to like
44:38uh
44:39be strong in in this situation where the
44:41rest of the organization like completely
44:44disagrees with the team. And then you
44:46somehow you have to create an
44:47environment where first of all, like
44:50we we believe that ideas get much better
44:52when they're challenged. So, you on one
44:54hand you really want to facilitate that
44:56type of really honest, even brutally
44:58honest feedback. But at the same time
45:00you have to be very clear that nobody,
45:03like nobody outside the team, has any
45:06authority over the game. It really is
45:07the team's decision, like whether you
45:10you like it or not. And and you know,
45:12like and it's interesting like uh
45:14this is the same case with almost all of
45:17our games. And even Brawl Stars, which
45:18is currently in beta, it's it's the
45:19latest example. So, lots of people at
45:21Supercell
45:23thought didn't like the controls, they
45:25didn't like the art style, they maybe
45:26even didn't like the game genre. They
45:28don't believe that this type of like
45:29twitchy action game even belongs to
45:31mobile. But yet, you know, the team like
45:35like really really believes believes in
45:38the in in the game. And of course, like
45:40it's according it's in our culture to
45:42like give them the possibility
45:44to to create the game, put it out and
45:46work with the players to to make it a
45:48success.
45:50And we really like we try to create this
45:52environment where you don't ever need to
45:54create the proposal about about
45:56anything. I mean, you can just do it.
45:59And and you know, we just think that
46:00that
46:01in our culture where nobody like
46:03hopefully creates any kind of internal
46:05proposals about anything. I mean, the
46:07the minute somebody like proposes to do
46:09a proposal, I think no, no, you don't I
46:10don't need to see it. I mean, here's my
46:11feedback and and you just do it and and
46:13prove that it works. And you know,
46:15sometimes it works, sometimes it
46:16doesn't. And if it doesn't, that's fine.
46:20Uh
46:21And and then like the the last thing I
46:23want to say about this that that it also
46:25gets harder in case of mistakes. So, say
46:27that there was a big failure or we do a
46:28big mistake. And you know, we do many
46:30failures. We do failures we fail all all
46:32the time. Uh and and then I I think it's
46:35as I said, it's an essential part of the
46:37creative process. But then the the
46:39natural human
46:41kind of instinct in case of failures and
46:43in mistakes is to but okay, let's not do
46:46this again. Let's create this rule or
46:49process that will prevent this from
46:50happening again. And and and we always
46:53have to remind ourselves that hey, like
46:56people like let's remember like the the
46:58kind of foundation of the company. Like
47:00instead of rules, we believe in trusting
47:03the people and their common sense. And
47:05and you know, like the more we grow,
47:06actually actually the less rules we
47:08should have as a as a company. But
47:10again, there's like this temptation
47:12always to to to take take that path.
47:18So, I want to close by by saying two two
47:21things really.
47:22The first one is that I want to
47:24emphasize that despite everything that I
47:26just said, I mean,
47:28culture has not created our our four hit
47:31games. I mean, all the credit it it
47:34really does belong to the teams behind
47:36these games. And and culture has been an
47:39enabler at at most. And it it really is
47:41all about the teams. And for example,
47:44the story about Clash Royale that I just
47:46told you, I mean, that would not have
47:49happened without those people, those
47:51amazing people who were part of the
47:53journey and contributed to the to the
47:55game. And again, the role of the culture
47:57is just to enable these people and
48:00enable these teams to do exactly that.
48:03And then of course, there's always that
48:05element of of luck that it's related to
48:07timing that's in in in the play.
48:10The other thing I wanted to say is that
48:12that like how we start how I personally
48:15start have start to think about culture
48:16is that it's almost like how I'd like to
48:18think about our games. So, I believe
48:22that our games should be getting better
48:24like every single month. And as I said,
48:26our dream is that people will play this
48:28play our games after years, if not after
48:30after decades. And and in the and and
48:33sort of in in a way our games are never
48:35complete. And in a very same way, I
48:37believe that culture is never complete
48:40either. Like it it can always evolve and
48:42it can and it should get better. And I
48:45think we are a good example. So, like
48:46our culture has evolved
48:48on on many different
48:50respects and I I do believe it it it it
48:52is getting better. It can and it has to
48:54get better over the over years.
48:57Okay,
48:59that was it. As I said in the beginning,
49:00I hope you got some new ideas, at least
49:03some of you got some new ideas out of
49:05this this talk. Thank you very much.
49:17I think we have time for for a few
49:18questions.
49:25Hello. Yes. Hi. Hi again, Ilkka.
49:28Um
49:29I had the privilege of working with
49:31Ilkka along with I mean
49:32at Digital Chocolate. He indeed gave his
49:35best to
49:37make it a good place and uh
49:39uh uh I think at Supercell you have
49:41innovated
49:43team structures in a different way
49:45compared to how we were operating at
49:46Digital Chocolate and you've seen
49:47excellent results. Congratulations on
49:49that.
49:50So, my question is at Digital Chocolate
49:52while you were making dozens of game,
49:55Supercell
49:56you made very few of them. So, why was
49:58the shift from many to few?
50:01Well, I I I I think like the thing that
50:03I've learned the very hard way over the
50:05years and and I think many people like
50:06who've been part of these companies is
50:08the importance of focus. I mean, games
50:11is such a competitive industry where so
50:13many smart people, so many smart
50:15companies that and and you know, we are
50:17a small company. So, you can only do a
50:18very small small amount of things if you
50:21if you want to do them at at the kind
50:23best of your abilities. So, it it just
50:26believe that it's it's about focus and
50:27and you know, focus and quality they
50:29just go hand in hand.
50:31Thank you. Thanks.
50:35Hi.
50:36Uh yes, so you talked a lot about you
50:39know, the structure in terms of the
50:42game designers and the developers. What
50:45about the ancillary people that we might
50:47have? Let's say the marketing teams or
50:49the technical infrastructure teams
50:52you know, the IT and stuff. So, is that
50:55shared among the teams or is that also
50:57you know, each team kind of has to deal
50:59with that? That's that's a great
51:01question and that's exactly that's an
51:02area where our I think our culture has
51:04evolved quite a bit. So, obviously like
51:05when we started there was a six of us
51:08and five of the six were game developers
51:11and again I was the only one who wasn't.
51:14And and and and you know, like the the
51:15the whole idea started from like
51:17thinking about game developers only. But
51:19then like then when the model like start
51:21to work out pretty well like like when
51:23we we thought that okay, why wouldn't we
51:24take this idea and apply it to some
51:26other teams as well. But it really
51:28depends like on the
51:30area of business how it works well. It's
51:32like
51:33and and it is important for us still
51:34have the game teams are at the
51:36at the center of the company and they
51:38sort of like set the pace for the
51:39company cuz we want the company to be
51:41kind of led by the creative people. But
51:44then like for example in the case of
51:45marketing like
51:47I would say that that our marketing
51:50teams they operate very independently,
51:52of course. And they're still trying to
51:53figure out like what's the best way for
51:55this team to be independent enough, but
51:57yet of course you need to link them
51:59somehow with the game teams because the
52:00marketing needs to kind of
52:03communicate the
52:06the vision of the game to the players so
52:07that it's actually in line with the game
52:09is is about. So, how do you kind of
52:11combine these two things and and and
52:13we've done all kinds of different
52:14experiments and I think we're we're
52:16still learning. But I would say that the
52:18goal always at Supercell Supercell is
52:20that we we we'd like to all of these
52:23teams to fall into this idea of like
52:25independence and
52:26and responsibility. But but there are
52:28like differences how how these teams
52:30have implemented it depending on which
52:33areas of the business they they are.
52:35Okay, thank you.
52:36I'm curious about the transition
52:39of the old structure to the new
52:40structure and
52:43I guess the the shift in mindset needed
52:45by everyone, but especially in managers.
52:48So, how did you deal with that and what
52:50did you do when people in management
52:52positions didn't want to lose control?
52:56Well, of course this is a this is a a a
52:59different company. Of course some of the
53:00same same people, but I I think it's a
53:04like
53:05combination of like like you know, just
53:07having more experience. The actually
53:09interesting thing is that I've actually
53:10spent I've still spent most of my career
53:13in the previous company. So, Sum and
53:15Digital Chocolate combined was 10 years
53:17of
53:17of my life and and now Supercell I think
53:19we're getting closer to what is 8 years
53:22now. So, I've been at less less time at
53:24at Supercell, but you know, like I
53:27I believe that those
53:28those years have just taught us so so
53:30much how to do things and and and maybe
53:33maybe also like how how not to do do
53:36some things. And and but you're right
53:38like the kind of be the the mental shift
53:41that has to happen is that you have to
53:42like let go. You have to be like you
53:45have to like completely trust the these
53:48people and and and these teams. And and
53:50and as I said it's an easy thing to talk
53:52about when things are going well and
53:54it's way harder where where when things
53:55aren't going that well. But you know,
53:57but I think there's like enough of
53:58critical mass of people at Supercell and
54:01and and definitely not just the
54:02management. We have a very small
54:04management team and and we don't even
54:06call it that that way, but but it
54:07there's so many
54:09Supercellians at the company who like
54:11seen how this model can work and I and I
54:13think we all the Supercellians do a
54:15great job of keeping each other like
54:17honest about it. So, if and and we we
54:19often say that that if you see something
54:21at Supercell that it's not in line with
54:23our values and culture, it's your
54:26responsibility to speak up and and trust
54:28me like people do speak up if there's if
54:31they see that something is wrong. Thank
54:33you.
54:34Yeah, hi. Uh
54:37being the least powerful CEO that like
54:39that is your aim. So, like what what
54:41forms your day like what is your
54:42day-to-day schedule? And like what are
54:45your
54:46you know, like what are what is your
54:47vision? Like what do you wish to do?
54:49Yeah, so what do you say that I actually
54:50do? That's that's funny. I actually was
54:52just asked about that question not not
54:54long ago.
54:55And I really had to think about it. So,
54:57what do you say that I do? But
54:59so, I I I try to spend most of my time
55:01with in the people related things and
55:03and uh
55:04uh so, I
55:06I still interview like every single new
55:09employee
55:10that comes to the company like no matter
55:11what is the position. Uh
55:14so, I spend time with with recruitment.
55:16I I I do spend time with with all the
55:18team leads. Uh
55:20And then like of course everything that
55:22is somehow
55:23related to culture like at some point it
55:25will end up in on my desk even though I
55:27I firmly believe that you know, but it's
55:29not it it obviously isn't not my culture
55:32and it's not my responsibility. It's
55:34everybody's culture and everybody is
55:35responsible for it. But I
55:37it's my job to sort of somehow like
55:39summarize it for everybody and figure
55:40out what the culture is and and kind of
55:42keep on like telling telling that
55:44message.
55:45And then there's like lots of like not
55:47like very sort of just boring
55:49administrative stuff. So, like for
55:51example like everything that is related
55:53to like everybody in the company like
55:54stuff like you know, compensations and
55:56all these things. I mean, thank god they
55:58don't do any annual performance reviews
56:00or anything like
56:01that you know, those used to take too
56:03much of our time and like in the
56:04previous company added like no value
56:06whatsoever. But so, we don't do that
56:08type of stuff.
56:09But you know, like there are certain
56:11things like for example the compensation
56:12review which is
56:14you know, very much like on on on my
56:16desk and which I think is important. Of
56:17course I feel it's very important to
56:19make sure that people are compensated
56:20fairly. It's combination of those type
56:23of things. Thank you.
56:25I'll I'll have
56:27Thank you for sharing the insight.
56:29Actually, I've been working with
56:31a very big Chinese game developer and uh
56:34Um so, this is more like a question
56:36let's say from our president. So, let's
56:38say we have a super independent super
56:41editing who can make all the decision
56:44and execution.
56:46And there is a way like say if they if
56:48we can do it here, we can do it
56:50anywhere. So, what do you think would be
56:52you know, the potential risk that the
56:55team might go like like really
56:57independent like over independent
56:59somehow. How I can like convince let's
57:02say our president to give us the you
57:04know, the the power you know, like being
57:07a cell?
57:08That's a great question. I mean, then
57:11a long time ago when we actually had to
57:13had to raise money and and had to be
57:15dealing with venture capitalists. But
57:16actually was always the question when
57:18they were doing that like how why don't
57:20these guys like just leave and you know,
57:22set up their own companies? And I told
57:23them that that if they do that I mean if
57:26they feel like that that you know,
57:27that's the if that's the best place for
57:29them to develop the best games, that is
57:31exactly what they should do. And you
57:32know, like the way I think about it is
57:34that it's just like we need to make sure
57:37that Supercell is the best place for
57:39these people and for these teams to
57:42to create the best games. And as long as
57:44we
57:46as long as we we are successful in that
57:48then you know, everything else will kind
57:50of take care of of themselves. And if if
57:52you're not successful then obviously
57:53these people will leave
57:55if the environment isn't great enough.
57:57So,
57:58but you know, we and we we we we do talk
58:00about this this this
58:02quite a quite a bit and and really like
58:05idea I'd like to
58:06think the way I think about the
58:07Supercell structure that ideally it
58:10would
58:10combine the best of two worlds. So, you
58:13can work in a very small independent
58:16almost like startup within a startup and
58:19you can do your own thing. But then of
58:20course you get sort of the job security
58:22and and some of the other benefits that
58:24are part of being a bigger company and
58:26and and you can
58:27you just have more resources. I mean, we
58:29can do
58:30the reason why I believe it's more fun
58:32to work at Supercell today than it was
58:35like 6 years ago is that today we can
58:36just you know, do much bigger things
58:39just you know, because of the resources
58:41that we have available for us and we can
58:42run after much bigger dreams and that's
58:44exciting for me.
58:46Thank you.
58:48Maybe we have time for one more.
58:50Yeah, can I ask?
58:52Um so, teams very important obviously.
58:55Could you talk about how you form them?
58:57How they change over time? Do they
58:59change once the game goes live? Like how
59:02do how you manage the composition of
59:04them?
59:05Uh so, it it it's really is a group
59:06effort. So, what we do is that
59:09and big part of my job is that I I meet
59:13let's limit the discussion to these
59:14these game teams. So, I I meet with the
59:16game leads. These are people who lead
59:18the game teams every single week. So,
59:21and and that's actually the only regular
59:22meeting that we have by the way. But we
59:24have this 1-hour meeting and we just
59:26meet usually like without any agenda and
59:28we just talk about the teams most of the
59:30time. We don't make any decisions
59:32regarding the games. We never talk talk
59:34about like I mean of course if somebody
59:36needs an advice they might raise up the
59:39question in that meeting but most of the
59:40most of the time we talk about the teams
59:42and you know like maybe a team needs
59:44help in a certain area and and can
59:46somebody like move from this team to
59:47that team and you know those type of
59:49things and
59:50and I wish there was like more science
59:52or a process to it I mean because that's
59:54the hardest part of our job is to create
59:57great teams and and ultimately the only
59:59way to really figure it out is
1:00:01whether a team is is is working is just
1:00:03to put it together and see how it works
1:00:04out and then if it doesn't then you can
1:00:06make changes. Cool. Thanks. All right.
1:00:08Okay. Thank you again.

About this transcript

This page contains the full transcript of The Cell Structure: How Supercell Turned the Traditional Org-Chart Upside Down by GDC Festival of Gaming, generated from the public captions YouTube serves with the video. The transcript has 12,132 words across 1,792 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.