The Washington PostDemocracy Dies in Darkness

Gabe Newell on Valve’s intimate relationship with its customers

Placeholder while article actions load

Valve is one of the most successful video game companies in the world. The firm's online game distribution and multi-player platform Steam has 65 million users. At this week's CES conference, the company will announce hardware partners for one of its most ambitious undertakings so far: a line of gaming console alternatives running the company's linux-based Steam OS.

In November, I sat down with Valve CEO and co-founder Gabe Newell in the gaming company's Bellevue office for a feature story.  The first half of the interview focused on the company's management structure. In the second half, Newell argues that the company's relationship with its community is key to bringing in value -- and touches on his own relationship with fans. This interview has been lightly edited for length and clarity.

Sony’s PlayStation 4 and Microsoft’s Xbox One might face new competition from Valve’s Steam Machines. At CES 2014 Valve unveiled 13 new consoles that make popular online games easier to play on a TV. (Video: Zoeann Murphy/The Washington Post)

Could you tell me about Valve's relationship with customers, and how that’s changed over time?

Well the Internet has really changed the relationship between customers and product or service providers. I think that what we represent is sort of the leading edge of something that will affect a lot of industries. The lines are a lot blurrier between who is a customer and who is a creator. It used to be it was pretty easy, right? You make cars, you send them out to some form of indirect distribution, and customers buy them, and you basically never hear from them again. Whereas in our world, sometimes people are producing as much as they are consuming in the experience. If you look at a multi-player game, it’s the people who are playing the game who are often more valuable than all of the animations and models and game logic that’s associated with it.

So, for us, it’s a lot easier to realize that the experiences we create for people are really a collaboration between all of the members of the community. This sounds a little bit touchy-feely granola, but it’s pretty concrete: When you have kids making six figure incomes making art assets inside the framework that you created and they’re able to sell digital goods and services around the world, it’s pretty concrete that you have to recognize that you are in a partnership with your customers. It’s always a collaboration.

Certainly game design is another thing that forces you to recognize that it’s much more of a dance between the actions and choices of a gamer and you as a creator. You can’t pretend that it’s just a canonical experience that lives in your head and is violated by being turned into bits, and a customer’s job is to approach it and try to understand the auteur’s vision of the experience. If you come to it with that sort of model, you’ll never ship a game. So the useful thing about having to recognize that you’re collaborating just to even ship the simplest kind of game is that it starts making you think about all of the other ways you should be collaborating with your customers.

What we've always found is that the closer we are to our customers, the more introspection we have into their choices and values, the more money we make. One of the biggest thing about Steam was that it was the tool that we needed in order to do that. I remember back in the early days of Microsoft that from the day that you decided that you were just going to put out an ad to a customer -- and all you were usually able to tell them was that a new product was available -- it was about nine months before you could actually reach the first customer. You were essentially throwing lawn darts into the future and hoping that you were going to make a decision. And you couldn't tell if you picked the right features to do, or if you'd been making good choices about how to prioritize stuff.

Now, an individual at Valve can basically get data back from 10 million people is maybe 10 minutes. Being able to have that for everything you do is insanely valuable. All of the sudden you're no longer dancing in the dark with oven mitts on -- you can actually see how your audience and your customer-base is reacting to what you've done, riffing on it and extending it.

And you just referenced your time at Microsoft. Can you tell me about your decision to leave Microsoft?

Mike Harrington and I were sort of the generation of people at Microsoft whose view of what the company was that Microsoft was the best place in the world to build software -- sort of the blue-collar factory way of thinking of it. There was a lot of craft knowledge to building software, and that's where everybody's pride was. We started before there was an Office or a Windows. So we learned a lot about how to make software, and we decided that because we knew the guys at id Software [makers of Doom and Quake], and we were like, "I think we could make game software. The things we've learned about making operating systems and application software seem to apply. It seems absurd for us to go into the entertainment business, but we at least know how to make software." So we went down and talked to the guys at id, and they were like, "here's the source code for Quake." And we were like, "Guess, we're going to start a games company."

You know, we thought about it, and we were trying to analyze whether or not it was an interesting business to be in -- because we'd both seen lots and lots of kinds of business, we both worked in the operating system business, and you're always talking to your partners. Some of them are United Airlines, some of them are a hundred thousand programs, some of them are one-person start-ups. The games business looked like one where you'd know pretty quickly if you were any good at it. It seemed like the big mistake would be to get into a business where you couldn't tell if you were any good at it because you could throw a lot of money away and find out that you really had nothing to add. So the games business looked attractive: It had a heavy software component, we had the support of an existing industry player, and we both thought that we would fail miserably and go back and beg our old bosses at Microsoft for our old jobs back. And then Half-Life came out, and we were like, "Wow, I guess we can actually do this."

We were really trying to think about what were businesses where you can differentiate. Ninety percent of games lose money, 10 percent make a lot of money. And there's a consistency around the competitive advantages you create, so if you can actually learn how to do the art, the design and the programming you would be consistently very profitable. And it seems crazy, but at the time it seemed to us that the future of entertainment was going to be interactive -- that we'd move away from passive forms of entertainment toward ones that would be participatory. We both thought we knew there would be fundamental drivers for why participatory experiences are more valuable to people than passive ones.

We looked at it, and we said, it has the potential for being a good business if you get good at it, you'll both be profitable, and the business will be growing further.  Plus, Mike and I both liked playing games.

I was going to ask -- you haven't mentioned here at any point that you really enjoyed playing video games...

The real story is that we were both rabid game players, and this was an excuse for us to play games all the time.

How do you define your company philosophy? Explain it to my like I'm five. 

Give people the tools to focus on making customers lives better, and throw away everything else. So, get rid of organizations, get rid of titles, put everybody's desks on wheels, and just say "here are the tools you have so you can tell whether you are doing a good job or not," and support people as much as possible with the colleagues and tools they need to do that.

I spoke pretty extensively with [Greg] Coomer and his people a few weeks ago, and one thing that stuck out was this idea that you're taking a lot of the same customer focused approaches Valve has used in software development to the hardware space -- particularly with the beta that you're doing, sending out the machines, soliciting feedback. 

Yeah, an example could be that we're going to workshop the controller. So in the same sense that customers can take our content as a starting point and build on it -- and they can build stuff that is valuable to specific members of the community, like some people really like pandas? So you want a panda courier? Somebody is going to make it for you.  So we're going to apply that. It's just a valuable in the hardware space as in the software spaces: People have different sized hands, different trade-offs and choices they want to make.

Don't think of it as a hardware or a software issue -- or a game or operating system problem. Think of it as a community of customers: How can you take these systems and build cycles out of them. It's great that we have our software as a way for people to add value to it, and the hardware should be the same way. We started off with giving people models of our characters, and now we're starting off giving people models of our circuit diagrams for our controllers and the Steam Machines.

They'll start innovating. They'll do a bunch of obvious stuff, they'll do a bunch of terrible stuff, and they'll do a bunch of amazing stuff that's hugely valuable and illustrates why the open community-oriented approach is the right one for a wide variety of post-Internet services and products.

Can you put your approach to this community aspect of business development in the context of how you think other major tech companies are working right now? 

There's a pretty important paper written by an economist named Coase called theory of the firm, and he said economists have this weird issue to deal with which is, "given markets, why do you need corporations? Why do they exist?" What is in the price -level, to use an economist's term, to properly allocate capital and properly organize labor.

And his answer was there are a set of issues related to discovery costs, transactional costs, and obscured reputation that meant that it was simply too expensive to keep doing these things over and over again, and corporations were solutions that solved that problem. When you looked at where we were in the 80s, it seemed really clear that at one point in time, distribution represented 50 percent of gross revenue. With hardware products, very steadily those margins declined. Businessland is the answer to a trivia question, not the dominant method of selling computers in the United States. So what you had to look at are what are the structural changes that were occurring, and it really was about the IT transformation and that the costs of moving information around and processing information was dropping exponentially. When you look at the choices that we made about our organizations, it's really about the huge reductions in cost about communications and information processing.

It's not that some kid in Kansas -- my favorite story of one of our workshop contributors -- it's not that he couldn't have done it, to create something of value for people around the world, it's just that it was thousands or millions of time too expensive for him to market, and communicate, and distribute that product to kids in Malaysia or kids in Russia. But now that the costs have plummeted relative to the costs of the production. It used to be 99 percent of your company were things other than building the things that were of value, and now those costs have plummeted so all of the sudden it changes the relative balance in the organization of your allocation of capital. It's just a consequence of what has been happening with information technology in general, right? It used to be that you needed a $500-million-a-year company in order to reach a worldwide audience of consumers. Now, all you need is a Steam account.

That changes a whole bunch of stuff. It's kind of a boring "gee, information processing changes a stuff" story, but it's going to have an impact on every single company. I think you already see examples of companies today that are outperformed by self-organizing activities on the Internet. In 1923 when he wrote the theory of the firm, I think Coase would have said, "oh, that makes total sense that that would have happened -- corporations are not an indisputable and fundamental element of how economies operate, they're just the result of one set of relative costs at one particular time and if you said those all change, he would have said the nature of how to optimally do it will change accordingly"

Right now it seems really clear that those costs are going to continue to drop more and more and things that take advantage of that relative change will be beneficial. I think that you'll see more and more one-to-one product design, where every product is customized for every customer, which sounds bizarre given the United States's history of mass production and mass manufacturing -- which were perceived as huge steps forward, and they were. But now we're coming up on a time where we'll see if you fax me a copy of your hands, I'll give you a controller that fits your hand perfectly.

That's what the future looks like more to me than "we can make one hundred million identical controllers." And it's because a whole bunch of things that used to be only economically feasible when they were done over and over again while recognizing that they were an inferior service will now get to the point where the cost of doing the unique product for a unique customer is going to be so low that the advantages of mass production will be dwarfed by the "I have a better product, not as cheap as if there were 100 billion products that were poorly suited to me, but the benefit that I get from something that is optimized for me is worth it."

So corporations that get in place systems to take advantage of that will win out over companies that are fighting the last battle to make the cheapest, identical widget over and over again.

You mentioned there one of your favorite examples of a kid in Kansas. I'm a Kansan, so I'd like to hear more about that story.

Well, he's a minor. He's a really great kid, but please don't write about him without checking with his parents. [Editor's Note: We did check with his parents before our original story, in which we spoke with him and his mother.] His name is Ben Henry. He was 15 when he started doing this. His parents are charming. We talked to them, and we said "he started off building items for Team Fortress 2, and he has this great artistic sense for what what people will like, and design sense."

In the sort of old world, he never would have even found a company to be a game developer at -- plus nobody would have been able to recognize how talented he was, he would have gotten some boring entry-level job and struggled for many years, and instead he's really good at it, and he's doing it. And he loves it, and he's making a ton of friends and collaborating, and his world has just opened up. He gets people sending him e-mail in Ukrainian and translates it using Google translate. He has friends and customers in foreign countries and foreign languages which 20 years ago never would have happened.

My experience about just how the world is different, upstairs, I'm listening to my kids shouting in Russian. I'm like "you guys don't know Russian!" But they were downstairs playing a multi-player game with and against Russians. We beat on them with sticks, and they refuse to learn languages in the classroom, but they were absolutely motivated to learn fragments of pidgin Russian when they were playing with and against Russians. I think mostly they were learning curse words, but they know way more Russian than I do, and they also know Japanese, and German, and Chinese -- and it never would occur to them that they are studying foreign languages, but they sure are picking it up. The world my kids are growing up in is profoundly different from the world I grew up in.

We've talked a lot about the company's relationship with customers. Could we change gears a bit, and talk about your customers' personal relationship with you... which seems to be very intense.

Yeah, I think that the intensity of that really is a reflection that we look like a post-Internet company to them. We sort of get the phenomenological world that they live in -- we act like we're on reddit, right? We get 4chan jokes. My e-mail is public. I think what they're really trying to say is that, "oh, you're on this other side of this divide with us, the customers?" as opposed to the CEO of General Motors who has an army of people around him to keep other people away from them and stuff like that. I really think that's not about me personally, or so much about the company so much as a recognition that these companies that are engaged with their communities are going to be really well received in the future by mass audiences.

We've also been around for a while, and I haven't pissed them off too much. Which they're grateful for because a lot of their relationships, or things that they're fans of, failed them at some point or another. We seem to be kind of reliable and trustworthy, and they like that.

And they hug me when they run into me. I'm not a hugging person, but that's what they want.

So really what you're telling me is that it's all about giving the customer what they want, even if it's hugs you are not entirely comfortable with. 

I was with my kids the first time that happened in public, and my kids were pretty cool with it. But I wasn't. "Dad, roll with it." Even now, I'm learning from our customers.

Do you have any sort of personal reaction to the way that your customers seem to rally around you as an individual? For instance, I have spent some time on subreddits that are devoted to you. There's a lot of fan art about you. There were a variety of birthday videos for you recently when you aged -- not that you're not constantly aging, but when you hit that yearly mark...

My response to that is to talk about my personal motivations. Some people are sort of achievement motivated or in direct competition. A lot of my rewards are secondary -- I like to do stuff with other people and feel like we are part of some joint success. So to me this is just a sign that thinking of myself and the company are part of this broader community is the right way to do it.

I like our customers. They like me. They like us. And that's where the validation of our business principles and the choices that we make happens. The Internet is the Internet -- if you are visible to the Internet they're going to playfully mess with that identity, but that's just the way the Internet works. It's entertaining. At least there isn't a LOL Gabe meme -- cats have a right to be super cranky.

Getting back to gaming and technology... Do you have any idea what you think is next? I've left that purposefully broad.

There are a lot of obvious things that are going to be happening. I think the big thing is just seeing those issues being worked through, and different people are going to see various opportunities in that and try to come up with novel ways to take advantage of it.

I really think that probably the most outlandish idea I have is that the Internet will replace portions of functions traditionally performed by corporations -- that it'll be better at those old Coasian concepts of organizing labor and allocating capital: That the Internet will beat corporations over the next 20 years. For me, I don't have to make a bet one way or the other on that, but intellectually, it seems consistent with everything I've seen happening. The forces at work will continue to make individuals the, sort of egalitarian and democratic view of how we produce stuff as an economy is the one that I see increasingly winning out.

Read the feature story based on this interview here. Part one of this interview is available here.