Lessons from Valve — How to build a designer’s paradise
This post was originally published on August 31, 2011, on my personal blog.
The building had somehow been lost in the skyline every time I had seen Bellevue. Half an hour before my interview, I walked into the lobby with my heart racing — it’s not every day that you land an chance to work at your dream company. A directory was posted by the elevator, but the company I was looking for wasn’t on it. As I searched the walls for some evidence I was in the right place, I heard two men behind me joking about a girl working at Cafe Ladro… and saw that one was wearing a Legend of Zelda T-shirt. I followed him into the elevator, waited for him to push a button, and pretended that I knew where I was going. Several floors later, he stepped out and tapped his keycard against a glass door, laughing at a joke his friend made as they walked in. The door closed behind them… marked Aperture Science.
Valve doesn’t exactly want to be found. They don’t have a guest store, there are no signs pointing to the entrance, their address is listed as a P.O. Box on their website. Try searching for them on Google Maps.
It makes sense. Valve is notorious for decade-long development cycles, groundbreaking games that change the definition of their genres, and absolute secrecy. If the development time of its famous Half-Life series holds to form, the fans won’t know how the story ends for more than twenty years after it began. They’re the butt of constant jokes, with a fanbase that’s as dedicated as Apple’s. My generation of gamers grew up playing Valve games, and their treatment of games as art has had a marked impact on the growth and breadth of the industry. If Valve were easy to find, fans would be breaking down the doors.
Their recruiting is equally enigmatic. Colleagues of mine far more talented than I mentioned how they had sent in resumes, but never heard back. Stories echo through the gaming community of Gabe Newell hiring entire teams of students after seeing Digipen projects. Not long after being offered the contract, I attended a gaming industry night at a local bar. A developer for XBOX, after hearing where I’d be working, told me, “you’ll never do better than that. Gaming industry giants only go one way, and that way is to Valve.” Internal documents include long target lists, where Valve employees add names of the best and brightest innovators in the industry, and if/when they will be available.
Valve hires the very best. They’ll wait for a year for the right candidate to become free instead of hiring someone who is merely great, and immediately available. “Once you’re here, you don’t leave,” one Valve designer told me. “No one leaves… at least to go elsewhere. I’ve known a couple who quit to start their own companies, but beyond that, people like it here.” There’s a reason for that — Gabe Newell knows how to make a designer’s paradise. Here’s how.
Establish a culture of design
If you’re a professional designer, your value is in your taste, and in your ability to replicate that taste. Both of these abilities require years of work, dedication, and study. Yet, while most clients are aware that it takes some degree of expertise to use Illustrator, they don’t always believe that it requires any sort of special knowledge to know what looks good and what doesn’t. Can’t everyone tell if something is ugly?
This causes all sorts of conflicts between client and designer. The client will have an idea of what the finished product will look like according to their own taste. What the designer creates may or may not be in line with that taste, which leaves the designer in a conundrum — deliver a shabby design that makes the client happy, or a great design that the client doesn’t want? Usually, the client wins, and the designer omits the product from her portfolio and rants on ClientsFromHell while continuing to pay rent.
Not at Valve.
A significant portion of the employees at Valve come from a design background, and know what this conflict is like. Interviews and hiring decisions are conducted by the team that would be acquiring the new hire, so new designers are hired by current designers — not by an HR department, and not by executives. They’re looking for someone with good design sense, well-rounded ability, and self-directedness, not necessarily for their tastes to align with the interviewee. If they hire you, it’s because they respect and trust your ability. You know when the product will need to be pushed live and you know the standard of design Valve has, so you know what needs to be done. In the case that a designer has a question or seeks feedback, the responses are more likely to be directional instead of specific: “we’d like to emphasize the TF2 hats as part of this message; could we put more focus on that?” instead of “could we try a different font/color?” The designer is given goals and clarification on those goals, but is never told how to achieve them.
Hire the best self-driven people, set your expectations, then set them free.
Specialization is Tyranny
Every desk at Valve is, in fact, a mobile workstation. The tower hangs under the desk in a metal sling, and all of the cords are plugged into your own power strip, which then plugs into a single outlet on the wall. The desk itself is on large, roller-blade style wheels, and there’s a large freight elevator in the middle of the building accessible to all. It’s an excellent representation of the Valve internal philosophy of hiring the best, then setting them free: “that’s why we put wheels on desks. People say they have to go work on this different problem now, because nobody else is and we need to get it done in order to get other things done. So they pack up their desks, and move to their relevant teams,” said Gabe Newell in an interview with Develop Online.
As I was finishing up my contract at Valve, I spoke with one of the lead designers for Steam about what he did, and how he got there. He had done print design, flash prototyping, HTML and CSS UI, and even some hardware hacking before he made the move to Valve. “We don’t hire someone who can only do one thing, even if they’re the very best at that one thing,” he said (I’m paraphrasing from memory).
If there is any single thing that creates the culture of focus on the product at Valve, this is it.
Compare two scenarios — one at Valve, one elsewhere:
Valve: a marketing and publicity person is writing some copy for an article for Portal 2 in a gaming magazine. He finishes it, then sends it out company-wide for a preview before submitting it to the magazine. A Steam back-end engineer responds with a suggestion on emphasis and word choice, and a Steam designer chimes in with an argument for consistency on review quotes. After a half-hour email chain, the marketing guy puts the engineer’s word choice in the article, and both designer and marketing guy implement a new review quote found by another guy working in merchandising.
Elsewhere: A company has decided on a new design for their website. The head of marketing wants to tell the story of the product on the landing page, and commissions an outside creative firm for a flash introduction. The firm creates the introduction, clocking in at over a minute long and 3MB. A hired consultant suggests combining the flash animation with further explanations of the product and its story, resulting in a final animation over two minutes long and 7MB. The marketing director agrees, and passes the order to implement the new introduction on the website. The web developer argues against a flash introduction at all, let alone a 7MB one, bolstering the argument with site traffic bounce rates. Considering the site content to be a marketing decision, the flash introduction remains.
“The first thing you should know here is that Gabe [Newell] is on top, and there are 249 people below him. That’s the whole hierarchy,” one of the Steam designers told me on my first day.
At Valve, there are no real titles. “The first thing you should know here is that Gabe [Newell] is on top, and there are 249 people below him. That’s the whole hierarchy,” one of the Steam designers told me on my first day. Titles specialize employees, and put them in a little plot where they’re allowed to work. Specialization sets employees against each other, carving out little kingdoms of responsibility. Anything outside their kingdom is beyond their purview, and anyone stepping within that kingdom is encroaching upon their pride and their job security.
At Valve, the job is the product. Everyone makes decisions, everyone invests in it, and everyone takes pride in the results. Breakthroughs in Valve products have come from completely unexpected sources, and from employees with specializations that the company didn’t even know about. Next time you play through a Valve game with audio commentary, pay attention to their stated job titles.
Hire the best well-rounded collaborators, set your philosophy to be product first, and let the work assign itself.
Creating a product? Permit creativity.
Before products are born, they’re conceived. They start as a spark, they gestate in the dark for a time, and then they’re born. The parents of that product have hopes and dreams for the future, and do everything they can to shape the product to be what they want it to be.
It’s a tired but apt metaphor. At some point, products take on a life of their own. As the world interacts with them, they’ll be used in ways its creators never dreamed of, weaknesses will be discovered, and it’ll have to adapt to survive.
When Valve started production on Portal 2, they gave their game developers six months to work on absolutely anything that could contribute to the development of the game. Several developers worked together developing a game concept they called “F-Stop," which used an entirely different gameplay mechanic from the Portals players were used to. It was well loved within Valve, but the fans of the game wanted to see the stories of Chell and GLaDOS continue, and also wanted to further explore the portal gaming mechanic. Significant portions of F-Stop concepts ended up being absorbed into Portal 2’s excellent Co-Op campaign, and game mechanic may yet see the light of day in a future game release. F-Stop was just one example of numerous projects and experiments that had a significant influence on the game, and it’s quite apparent in the depth and feel of Portal 2. Several revolutionary features are on their way to Steam as we speak, all incorporated by side projects from the Steam team. Get ready to hear about “big picture mode.”
Valve isn’t the only one to recognize the value of unleashing its creators. Google famously allows their employees to devote 20% of their time to whatever they desire that may add value to Google’s products. It’s a bold move, but one that has paid off time and time again for them. This turns out to be an incredibly difficult thing to do — it’s contrary to rapid development cycles, firm release dates, and our very psychology. It seems, however, that taking those risks contribute to Valve’s and Google’s edge over their competitors.
Hire the best creators, then step back and give them the freedom (and the time) to create.
The Valve philosophy is actually a very simple one. They hire the very best, step out of their way, and let them create according their own best judgement and the voices of their users. It’s a formula that has led Valve to be one of the most respected and revered creative companies in the world.
