THE BIG PIVOT

Stewart Butterfield,
Co-founder & CEO of Slack

Cameo appearances: Tim Lefler (Slack)

Photography By Christopher Schodt

About this episode:

Slack's Stewart Butterfield knows: You CAN pivot from failure to success. But only if you slash and burn everything that isn’t working. And then bring everyone else along. There’s more to a pivot than a sharp left turn. Transcript below.

"Concentrate on the storytelling, on convincing people. If you can't do that, it doesn't matter how good the product is."

- STEWART BUTTERFIELD

"What do you do with an idea that isn't working? You kill it."

- REID HOFFMAN

LISTEN TO THIS EPISODE NOW

< RETURN TO FULL EPISODE MENU

GUESTS ON THIS EPISODE

Stewart Butterfield

Stewart Butterfield is the co-founder and CEO of Slack, the overwhelmingly popular team messaging application. Before that, he was the co-founder and CEO of Flickr, which grew out of his initial work with social game networks.

 

TIM LEFLER

Tim Lefler has worked with Stewart since the days of Game Neverending. They now work together at Slack.

“This is the greatest software development methodology: Don’t think about what you’re doing, have no ego. There’s no speculation.”

-Stewart Butterfield

EPISODE TRANSCRIPT

REID HOFFMAN: It’s 2002—dark days in Silicon Valley. The dot-com bubble had burst. Tech investors were quiet. Almost nothing of note was launching online. And then out of nowhere came an experimental video game called Game Neverending—a role-playing game that was all about social interaction. You work with other players online to create a world. You build buildings, houses, and other objects. For the gaming nerds out there, it was kind of a precursor to Minecraft. You could pool your resources with friends to build structures, invent new objects, and reach higher levels together. Your objective? Just keep building; don’t stop. There was no way to win this game—hence the title, Game Neverending—but a goal that most hard-core fans aspired to was to build the final item of the game: a Game Neverending. So meta!

It was an offbeat, self-aware game that didn’t take itself too seriously. The prototype had a great cult following, with thousands of users who loved it. But—

 

STEWART BUTTERFIELD: But we couldn’t raise any money for that.

 

HOFFMAN: That’s Stewart Butterfield. You might know him as the founder of Slack. But before that, he created this daring new game—just at the wrong time. As I said, it was 2002—the dot-com bubble had just burst—dark days in Silicon Valley. And tech investors weren’t thinking about gaming as an online experience yet.

 

BUTTERFIELD: So it was a really black and bleak-looking point in the history of financial markets generally, but anything as frivolous as a game is just not going to get funded.

We got to the point where the only person on the team who got paid was the one person on the team who had kids, at that point. And we needed some kind of “Hail Mary.”

 

HOFFMAN: Ludicorp was running out of money. The company was failing. Having no idea what the next step was, or what to tell his team, Stewart needed to stall. So he and his co-founder Caterina Fake got on a plane to New York to attend a video game conference. But before the plane landed, things got even worse.

 

BUTTERFIELD: And this sounds almost made up. But on the way, I got food poisoning on the plane, puking on the Van Wyck on the way into New York, arriving at the hotel. Just being sick all night—it’s like 3:00 or 4:00 in the morning—trying to keep anything down, like ginger ale or water. And the whole idea for Flickr came to me.

 

HOFFMAN: Flickr—one of the most successful photo-sharing sites of all time, an idea that forever changed the way we interact with photos online. You could argue it changed the way that we interact online, period. And the idea hits him while he’s crouched over a toilet in a hotel with food poisoning—Silicon Valley lore at its best.

But at that moment, Flickr wasn’t exactly a game-changing vision of the future. It was a way out of the mess his company was in.

 

BUTTERFIELD: It wasn’t coming from a grand vision of what photos could be, All that stuff came later. There was no insight, it was just like, “Can we not go out of business?”

 

HOFFMAN: “Can we not go out of business?” These are the words that launched a thousand pivots. In Stewart’s case, it opened his eyes to a service that would do far more than change his online game—it would change the way we interact with photos forever. It would pave the way for Facebook, Pinterest, Instagram.

 

The idea had instant resonance. It caught fire with users. It got attention from investors.

 

The only problem? He still had this whole gaming company still operating. What do you do with an idea that isn’t working? You kill it.

 

I believe the moment you find the thing that your users cannot stop doing, you need to slash and burn the rest of your business.

 

[THEME MUSIC]

 

HOFFMAN: This is Masters of Scale. I’m Reid Hoffman, co-founder of LinkedIn, partner at Greylock, and your host.

 

I believe that the moment you find the thing that your users cannot stop doing, you need to slash and burn the rest of your business.

 

In Silicon Valley, entrepreneurs tend to celebrate this kind of daring pivot. They see an opportunity. They act—and they don’t look back. Later on, they sound a bit like Caesar reporting to the Roman Senate: “I came, I saw, I conquered.”

 

But the truth is a lot messier.

 

There’s more to a pivot than a sharp left turn. First, there’s the opportunity you’re pivoting toward. Can you see it clearly enough to navigate toward it? Can you convince others to come along?

 

Then you have to pivot away from your old idea. This can be incredibly difficult, because it involves humans. And humans don’t tend to let go of old ideas easily. You risk blowback from your co-founders, your staff, your investors and your users.

 

This will likely be the single greatest test of your leadership skills, because your credibility will come under scrutiny. Are you even believable anymore.

 

I wanted to talk to Stewart Butterfield about this theory, because he pulled off two uncanny pivots. He has twice launched game companies, only to pivot toward game-changing communication platforms: first, as the co-founder of Flickr, the pioneer in photo-sharing, and second, as the co-founder and CEO of Slack—a platform for group communication within teams that has taken off in Silicon Valley and beyond. Both companies, believe it or not, started as online video games.

 

And at first, this might seem like a real leap. But when you learn a bit about Stewart, it starts to make a lot more sense. For one thing, Stewart likes video games. But he doesn’t like-like them.

 

BUTTERFIELD: I like playing games, but I wouldn’t classify myself as a gamer. I don’t own a console. I definitely have played my fair share of Civilization, and there are stupid iPhone games that I’m embarrassed to say that I enjoy.

 

HOFFMAN: He loves board games and card games—old-fashioned games that bring people together. And he wanted to bring that same sense of a communal experience to video games. And one thing you should know about Stewart: He has a lot of experience with communal experiences.

 

Let me start at the beginning.

 

Stewart grew up in a quirky Canadian fishing village called Lund. Drive west along Canada’s Highway 101, as far as the road will take you, and you’ll arrive. There’s a welcome sign at the entrance to the town that reads, “Lund, the end of Highway 101. It also reads, “Lund, the start of Highway 101.” That’s your first clue that the residents of Lund are an open-minded bunch.

 

Lund was settled primarily by American hippies in the 1960’s—and that’s where Stewart Butterfield was born. His birth name was Dharma. He changed it to Stewart at age 12. As far as I can tell, Stewart has taken to the capitalist lifestyle swimmingly. But I had to ask him about those formative years.

 

HOFFMAN: So you lived with your parents in a log cabin that had no running water for a few years. Did this shape you as an entrepreneur?

 

BUTTERFIELD: I’d like to say that some of that “no running water, no electricity” rubbed off in me in an Abe Lincoln kind of way. But I don’t know if that’s actually true.

 

HOFFMAN: So he might not be an Abe Lincoln. But those years in the cabin gave him a particular lens on the world. It shaped, among other things, how he thinks about games, and their ability to bring people together.

 

BUTTERFIELD: My dad loved to play bridge, and he didn’t like playing bridge against the computer—there was just nothing there. He also wouldn’t have invited the same three people over to his house just to hang out and do nothing. But inviting them over to play bridge, and there was this magic. It’s exercising a part of your brain which is fun to exercise, but there’s also camaraderie, and there’s trash talking, and there’s a good amount of competition.

 

HOFFMAN: So when Stewart thinks about games, he’s thinking about something much bigger and more essential.

 

BUTTERFIELD: it’s not games that are so interesting to me, it’s play as an excuse to interact with people socially.

 

HOFFMAN: And this is the right way to think about Stewart’s first venture—Game Neverending. It was, at its essence, an excuse to interact with people socially. But it had some fatal flaws, among them: bad timing.

 

BUTTERFIELD: We were never going to raise money for the game. I had tried everything. I had put all of my savings into it, we had tapped out friends and family. We had more or less worked through all of the very small amount of angel investment we were able to get.

 

HOFFMAN: Stewart’s failing game needed a lifeline, and that’s typically the moment just before the pivot. There are some examples of peacetime pivots driven entirely by opportunity. But that’s not typical. Pivots almost always spring from failure. It was the failure of Game Neverending that forced Stewart to think about what would come next.

 

And his moment of revelation—over a hotel toilet in New York City—gave him the idea for Flickr, the photo-sharing site. But the idea didn’t come to him fully formed: it started out hazy, as many pivots do. The initial idea incorporated photo-sharing into the game itself.

 

BUTTERFIELD: We had this game interface. In the game, you had an inventory, you could pick up objects. We made that inventory a shoebox full of photos. And you could do interesting things, like drag photos around on to group conversations, and they would pop up on the other person’s screen. You could annotate them in real time. And there was chatting in the game, so you could talk to the fellow players—that became a cornerstone of Flickr.

 

Here’s the twist, though—that was the first idea for Flickr, which was actually really a terrible idea that had a lot of technological innovation, and so it wowed a lot of people, but it wasn’t a very useful product.

 

HOFFMAN: Flickr may not have been entirely useful—nor fully formed—when the idea first came to him. But the idea of sharing photos had immediate traction.

 

BUTTERFIELD: Maybe three months in, it was pretty obvious that Flickr had legs. So people started to use it, people were talking about it, it got a lot of good press.

 

HOFFMAN: For Stewart, Flickr was a sort of homecoming to his deep-seated interest in just hanging out. And that’s great for Stewart—but what about the rest of his team? They had signed up to build Game Neverending, and that kind of pivot can be brutal. I asked Stewart how they took the news that they were no longer a gaming company.

 

HOFFMAN: What was that turn like?

 

BUTTERFIELD: It was rough. And there was definitely differing opinions. There was a lot of arguments about what we should be doing. And for a time, we were working both on a game and on Flickr. The team was still really split, so we had a vote. And I remember this really distinctly, because I had to do some backroom politicking. After the first vote ended, it was a tie game. So I called up one of my co-founders, who was in New York, and had to do some horse trading. I felt like I was a senator.

 

HOFFMAN: When your team is split, it might seem smart to hedge your bets and pursue both ideas at once. It may seem to keep the peace with your team. But as a founder, you never want to say “We’re working on X, and we’re also working on Y—because my team likes both ideas.” That may be the expedient solution. But I can tell you how that story ends: like Thelma and Louise, holding hands and driving over a cliff together.

You actually owe it to your team to force a decision—X or Y, choose one. It’s ultimately the founder’s responsibility to make that call. But you have to bring your team along. This is the situation Stewart found himself in.

He almost had the votes he needed, but there was one hold out.

 

BUTTERFIELD: Well, I mentioned earlier that there was one person on the team who had kids that was continuing to get paid—that was him.

 

HOFFMAN: And this happened to be one of his co-founders at Ludicorp. Everyone else was willing to work without pay, but he had a family to provide for—mouths to feed, soccer camp to pay for. And this was the holdout vote that Stewart needed to stop working on the game, and fully focus on Flickr.

 

BUTTERFIELD: So it was a bit of a more, “Let’s play this out a couple of months, and see. If you’d like to continue to get paid, there’s only one way we’re going to do this.” And I didn’t mean that to be threatening at all, it just—we were never going to raise money for the game.

 

HOFFMAN: Hold on. Rewind.

 

BUTTERFIELD: “If you’d like to continue to get paid, there’s only one way we’re going to do this.”

 

HOFFMAN: Did Stewart just threaten that guy’s paycheck? Not exactly. You have to understand that Stewart had the kind of trusted relationship with his co-founder where this was permitted.

 

Working for a startup is kind of like going to war together. When you’re crouched in the bunker with the rest of your platoon, you form massive trust. If I’ve seen you look out for me, I’m going to look out for you.

 

So Stewart’s campaigning was pragmatic, rather than antagonistic. He’s saying “Look, we all love working on the game. But the reality is that there won’t be any money left if we keep doing this.” His co-founder heard him and got on board.

 

BUTTERFIELD: We had another vote the next day, and this time the Flickr side won. At that point, I think we had less to lose and there was less invested. It wasn’t as traumatic as the next big pivot, which was Glitch to Slack—that was seven years later.

 

HOFFMAN: And we will come to that.

 

HOFFMAN: Indeed, we will come to that. And you Slack fans may be eager for me to pivot. But this transition from Game Neverending to Flickr is just such a perfect example of an extreme pivot—I had to test a few more theories.

 

I believe that, as CEO, you always have to bring the core team along during a pivot. You have to make them feel like it was a joint decision. It doesn’t have to be completely democratic; in fact, it shouldn’t be democratic. But it has to feel democratic. People have to feel they have a voice. And as Stewart campaigned, he was essentially hacking the quasi-democracy, saying “Yes, your vote counts. But I can persuade you, because I have your interests at heart.”

 

Finally, the team could focus on Flickr, and slash and burn the rest. But what about your users?

 

BUTTERFIELD: There’s an easy part, which was we all liked each other, and we all liked making software, and it’s fun to make software, so we can just continue doing that. The hard part was just the psychology of sunk costs. We had written a whole bunch of great material, we had created this whole world, we had spent many thousands of hours convincing people to come play it.

 

We had built up this community, they were very enthusiastic. And every death of an online community is its own kind of tragedy, because those people think they will be able to maintain those relationships, but if the community is gone, they often aren’t.

 

That’s really the hard part, and that’s the part that is difficult for people to get over—even at the point where it was apparent that continuing wouldn’t ever result in something that was going to be economically viable, or even sustainable.

 

HOFFMAN: Game Neverending was never going to be successful. But a lot of people really loved it. And when the game died, the community died with it. To those players—those people—that was a tragedy. What’s most important to humans is their tribal membership, their relationships. And all of the sudden, there’s a pivot riding into town and saying “Hey, by the way, I know you’re all really happy being neighbors—but we’re shutting down the town.”

 

So Stewart and his co-founders have navigated what was in many ways a prototypical pivot. They’re going down a path, which was getting narrower and rougher, and seemingly leading nowhere. They see a different path leading to a better place, and they pivot toward it. They lead a team through a difficult organizational transition; they weather a storm from existing customers.

 

It might seem like the most harrowing transition was behind them. But keep in mind Stewart wasn’t just suggesting a hairpin turn for his team. He was suggesting a hairpin turn for the Internet.

 

BUTTERFIELD: Almost all the photo sharing at that time was entirely centered around printing, and that was the idea of what people wanted to do with their general photographs.

 

So that in one sense made it very easy. There was a lot of competition at the time—there was Snapfish, Shutterfly, there was Ofoto, which Kodak had bought, and Kodak had put a lot of muscle behind it.

 

And none of them thought of the Internet, as far as I could tell, as a social medium in itself, and a worthwhile place to put photos.

 

HOFFMAN: With Flickr, the sharing itself became the primary goal—not a print, or a mug. And if this concept seems awfully familiar, it’s because Flickr laid the foundation for everything we do today on Facebook, Instagram, and Twitter: tagging, sharing, following, memes. It was arguably the first comprehensive experiment in modern social media—a testbed for innovations.

 

BUTTERFIELD: Gradually, over time, we layered on the photos having titles, and descriptions. People could comment on them. Joshua Schachter from Delicious convinced me that tagging photos would be a good idea, and so we added tags.

 

And soon, we had a place where people could really play with them. I remember the very first tag for what just happened—someone posted a photo of a street lamp in the dark, and it was glowing in the mist. And they tagged it “glow project,” and they said, “Hey, other people, why don’t you upload things that are glowing and tag it with ‘glow project?'” And that was something that, as far as I could tell, had never happened on the Internet before—that invitation for joint creative exploration based on tags.

 

HOFFMAN: So if you’ve ever taken a “#selfie” or used “#foodporn,” now you know: It started here, with the glow project.

 

Flickr changed the paradigm of social interaction online. It succeeded because they had the right idea at the right time, and they made the right bets. But they also had another secret weapon: Good old fashioned manners. They knew that for Flickr to succeed as an online community, it had to be a community—period.

 

BUTTERFIELD: I think one of the strengths of Flickr was a really unique culture, and a huge amount of effort went into setting that. So Caterina Fake, one of the founders of Flickr, and my ex, she greeted every single user individually. There’s another designer on our team, a woman named George Oates, same thing.

 

And so I think a lot of what made Flickr great in the end came from that germ of a really strong and positive community.

 

HOFFMAN: We’ll hear more about building a strong community later this season—from Caterina, Stewart’s Co-Founder, who was equally fundamental to Flickr’s success.

 

The Flickr community scaled fast. And in 2005, Stewart and his co-founders sold Flickr to Yahoo.

 

BUTTERFIELD: It all happened so fast. We launched in February 2004, and by March 2005, the acquisition by Yahoo! had closed. Basically a year in market, and maybe 10-and-a-half months before we were in the middle of negotiations.

 

HOFFMAN: In 2008, Stewart left Yahoo. As an aside, he wrote an epic—and rather confusing—resignation letter to his boss at the time, Brad Garlinghouse. It’s odd and wonderful. Look for it.

 

Now, Stewart had money, connections, and a track record. He decided to launch a new project. You’ll never guess what it was—yes, another online social game. This one was called Glitch.

 

BUTTERFIELD: The idea was pretty much the same as Game Neverending, except that the technology was much better, so people could be much more creative and you’d be higher fidelity.

 

A game world with no combat, very absurdist—probably too high-brow, and too intellectual—Dr. Seuss meets Monty Python meets Borges. [Laughs] Alread,y you can tell that in an era where Farmville was just beginning to take off, this was not an easy sell to America’s casual gamers.

 

HOFFMAN: In retrospect, perhaps not an easy sell. But at the time, they were sure this game would be different. This game would be huge.

 

BUTTERFIELD: From our perspective, now we have some money, so that makes it easier; now we have connections. But also, computer hardware got, I don’t know, 100-fold better in those seven-or-eight years.

 

But now—now being 2008—there were tens of millions of people online in the US, and it became a real hotbed. It wasn’t something that seemed strange, it was something that seemed really natural. Computers were also faster. There’s all these new open-source technologies that we could take advantage of. We were more experienced and capable as engineers and designers. So the thought was, “Oh, we can’t fail this time.” There’s just no way, with all of these 10Xs that combine to make it 1,000 or 10,000 times easier for us. It would be impossible to fail.

 

HOFFMAN: That’s usually the Ozymandian warning—that thought.

 

HOFFMAN: So it seemed like the Glitch team had everything they needed to succeed, and the stakes were higher this time. Game Neverending had a small scrappy team with no money. Glitch had dozens of employees and over $17 million dollars in funding. It just had to work.

 

BUTTERFIELD: There were close to four years of development that went into it, and the first year was great. It’s kind of nice to make software when you don’t have any users—of course, I would much rather have many people using whatever I’m working on.

 

But if there are none, whew, it’s easy, and it’s fast, and we’re developing all this cool server technology.—stuff that’s just intellectually fun, even if it doesn’t actually turn into a product.

 

HOFFMAN: Stewart and his team were in deep—four years of development, 45 employees, tens of thousands of players, $17-and-a-half million dollars raised. Glitch had a small, committed fan base. But it was failing.

 

BUTTERFIELD: At this point, I’ve learned that this is not going to be a commercially viable creation.

 

It was incredibly powerful for the small minority of people for whom it worked. So they spent 20 hours a week playing, and they were very committed, it was a super-tight community. But most people, like 97% percent who signed up, would be out of there within five minutes.

 

The thing that killed it was just leaky bucket—I mean, just like very classic, you can put it into Excel in five minutes, and see that this is not going to work.

 

HOFFMAN: Glitch was losing money fast. For the second time in his career, Stewart needed a lifeline. With the benefit of hindsight, we know now how the story ends. And it would be easy for him to rewrite history, and make the line from Glitch to Slack a straight one. To his credit, he doesn’t. Stewart is brutally honest about the desperate attempts to find their way.

 

BUTTERFIELD: The last 18 months were just a series of experiments in like, what if we did this, what if we tried that? Maybe it’s just missing this feature. We were pretty good at getting people’s attention, because we had more of a reputation, we were more connected, so we could get lots of articles written about us.

 

It did always seem like the next thing was going to be what saved it. And by the end, I actually think it was a much, much better and interesting product than it was when we first launched it to the world, and maybe if we’d gotten there—

 

Something that just didn’t work.

 

HOFFMAN: You can hear in Stewart’s voice: he still asks himself whether he could have saved it. And these questions form the inner dialogue of most entrepreneurs in the thick of a transition.

Later on, successful entrepreneurs often tell the story of the one daring pivot. They leave out the fumbles and false starts along the way. The truth is, most entrepreneurs pivot many times before they find their footing. Sometimes you find your way cleanly to success. And sometimes, you don’t.

 

BUTTERFIELD: There was this night where I just lost faith. I just realized, OK, I had tried the 15 stack-ranked best ideas we could possibly come up with to turn this around, and I don’t think the 16th was going to work if the first 15 didn’t.

 

HOFFMAN: Stewart was facing a brutal moment of truth: He ran out of options even before he ran out of money. And he had the maturity to recognize it. What I tell entrepreneurs as they begin their pivots: stack rank your proposed solutions, from the most promising to the least. Start with your best few ideas. If they work, great. If not, move on to the next-best ideas. But often you’ll find your next few ideas are substantially worse than your earlier ideas. And that’s when you should say, “OK, I’m even less confident in these ideas. And that means I’m heading off a cliff.”

 

Part of being an experienced entrepreneur is realizing, “I have to do something different. I close the company down, or generate a whole new class of ideas.” In short, get on with the next thing. You never want to pivot from a bad bet to a worse one.

 

Stewart knew what the numbers were telling him. For the second time in his life, he would have to close up shop on a product and team he loved.

 

BUTTERFIELD: Man, it is so hard—because the job of a CEO is often just to come up with a story that enough people believe that you can make something happen in the world. You have to convince investors, and you have to convince the press, and you have to convince potential employees, and you have to convince customers. And I had done a lot of convincing of people, a lot of convincing of people to come work on this project, to leave whatever thing they were working on before, quit their job, get paid poorly in exchange for equity in something that just didn’t work.

 

HOFFMAN: Stewart pulled the plug on Glitch before the money ran out. But he found a series of graceful exits.

 

BUTTERFIELD: And that was a tough conversation with our board and our investors, because they actually wanted us to keep going. It was a tough conversation with the other co-founders, who also wanted to keep going.

 

We had those conversations, and then we called an all-hands. And I walk into the room—and it was an unusual all-hands; we hadn’t announced it before, so there was already some nervousness in the room. And I stand up, and start, and I didn’t even get to the first sentence and I was starting to cry. I locked eyes with one employee, who just three months ago had moved to Vancouver to come work with us.

 

He had moved away from where his in-laws lived, and his in-laws helped take care of his—I think at the time 18-month-old, or two-year-old daughter—bought a house, and now I was telling him they didn’t have a job any more.

 

HOFFMAN: That employee he locked eyes with? That was Tim Lefler, one of the engineers at Glitch. We asked Tim how he remembered that moment.

 

TIM LEFLER: Stewart came in, and had a very somber look on his face. He had puffy eyes; he looked tired. And we thought it didn’t look great, but it was Monday morning, so who knows? And he came in, and the very first thing he said to us was, “It’s over,” and that we were shutting the game down.

And a lot of us obviously knew what that meant. We were not going to be working there any longer.

 

HOFFMAN: One of the things I admire in this story isn’t just that Stewart took responsibility; he also took action.

 

LEFLER: Stewart and some of the web developers decided to build a page on Glitch.com.

 

BUTTERFIELD: We made a website called Hire A Genius, and we put everyone’s LinkedIn up there, and everyone’s portfolio, and everyone’s photo,

 

LEFLER: And the timing of everything was very deliberate, so that when all of the news agencies received the press release, there was always this bit of information that would be sent out saying, “And these people are looking for work, and here you go.”

 

BUTTERFIELD: And we worked on writing reference letters and doing resume coaching.

 

LEFLER: And Stewart and the group decided, we’re gonna keep working until everyone else has another job. I received many calls and emails from people asking for me to come in for an interview. After about a week, I had a couple of offers.

 

BUTTERFIELD: And we ended up getting every single person a job. We were able to give customers a choice of their money back, or they could let us keep it, or we could donate it to a set number of charities. So we were able to do that in a way that was really kind of elegant, and built a lot of good will that would be useful for us later.

 

HOFFMAN: Here’s the key to managing through a pivot— those employees, if they feel taken care of, they’ll keep taking care of you. Tim and his teammates threw an epic goodbye party for Glitch’s online community.

 

LEFLER: There was so little ill will that a lot of us, myself included, we still worked for another couple of weeks, because we believed in the product. We believed in the game. And so what a lot of us at Glitch wanted to do, is we wanted to kind of give them a nice send-off.

 

We had a horseman character. He was supposed to be a horseman of the apocalypse, and he was carrying this doomsday bell, and he would just walk around saying like the end was nigh, or the end was neigh. I’d imagine that seems like a joke that we would have put in it. There was a rallying cry from a lot of the users around, having these “end of world” parties. There would be a big party in this one area of the world, and everyone would hang out and jump around, and have a big social gathering within this world. And then it would get shut down, and then we’d move somewhere else.

 

The horseman had to be designed, the horseman had to be drawn and animated, had to have sounds attached to it—there was a bunch of people involved. So yeah, there was effort put in. Obviously, it was limited effort, because we were all busy trying to find work, but there was effort nonetheless.

 

HOFFMAN: Stewart and his team embodied this idea of ending in the best way possible. And this isn’t just the ethical, human thing to do; it’s also the smart, strategic thing to do. Those employees, those investors, those users—they’ll come back.

 

BUTTERFIELD: So it was really, really tough. And it’s easy to look back on it now, and it’s character building—and in fact, we ended up hiring that same guy back, and he’s been a Slack engineer from the early days—everything worked out well there.

 

HOFFMAN: While they were essentially shutting the company down, Stewart and his team start the “Hail Mary” process again—a process that Stewart is very familiar with by this point. They had money in the bank, and a mandate to make something, anything that could scale. They took a long look at all the technologies they developed for the game, and asked themselves: “Was there anything there?”

 

BUTTERFIELD: We definitely knew that we wanted to keep working together, and we still had money left. Investors didn’t want their money back, because—I don’t need to tell you, but why would you rather take a two-third loss when something might come out of it?

 

We still had five-and-a-half million dollars, or something like that left in the bank. So we didn’t have to shut down, and I think that there’s a real temptation to go to the last dollar and hope that there’s some kind of “Hail Mary” that will save it. The good thing about shutting it down early was that we had the ability to just not do anything for a couple of months.

 

Within a couple weeks—I think three weeks—we decided—and there was a lot of stupid ideas first—but we decided that this internal communication system we had developed while we were working on the game—that could be a product, and we started building that. And we had a great practice, but it was a vision that might have well been passed down on to the mountaintop.

 

HOFFMAN: So, Stewart didn’t get food poisoning on a plane this time, but once again, he found an unexpected way to pivot. He found his scalable product in a communication tool his team had developed for internal use—a chat-based tool for his remote team that would eventually become a product called Slack.

 

And at first glance, this goes against something I know to be true: The most successful pivots stay somewhat close to the company’s original mission. This is where Stewart is the exception to the rule. I typically tell founders to ask themselves: “OK, in the path we’ve taken, is there anything that we’ve seen or found along our path that could lead to an alternative play?” It’s not the kind of thing where you pull ideas out of nowhere. “Hey, we’re just going back to the whiteboard here, what do you think? Should we make an ecommerce site, how about a mobile game? What about a pizza delivery restaurant?”

 

No, no. “From our path, what do we see?”

 

A chat-based communications tool is a really far cry from the online game that Stewart set out to build. But then again, it was a tool they already developed and used themselves. So it gets at the more fundamental rule of successful pivots—there has to be some kind of compelling reason why this team is going in this direction.

 

And Stewart’s team knew exactly why companies would adopt this new tool—they created it to meet their own needs, to communicate quickly, transparently, and permanently.

 

BUTTERFIELD: When we first got started, and there was only four of us, the natural thing for us to use was a very old Internet protocol called IRC—an Internet relay chat. It pre-dates the Web by a couple years. And because it’s so old, it misses a bunch of features that are now considered just standard.

 

So one of the things it misses is store and forward of messages. So if I want to send you a message, and you’re not connected at the moment I want to send it, I just can’t. There’s just no way for me to do it. And so we built a bot that would log all messages that were sent when you were offline, so you could read them when you got back online. And once we had that, we were like, oh. Wow. It would be super convenient to be able to search these messages, and since we already had them in a database, it was easy to search them.

 

It wasn’t something that we thought about, or talked about. It didn’t have a name. It was just in the background. But as we started hiring people, they would join and they would have this archive of everything that happened before they got there—as opposed to most companies, where you hire someone and they start with an empty inbox. They’re completely cut off from that history.

 

But again, it was—and this, by the way, I think is the greatest software development methodology that’s impossible to replicate. Which is, don’t think about what you’re doing, have no ego. There’s no speculation, there’s no, “I can imagine a user would want to.” To spend a minimum number of minutes addressing the most aggravating problems that you have, and just use it, and then see.

 

HOFFMAN: So, as Stewart puts it, he had the idea that was “passed down from the mountaintop.” The core technology was already built. Now all he had to do was convince his team to make the product. Stewart could see three steps ahead at this point.

 

BUTTERFIELD: There’s this deck that I put together that was like, “I think we should make Slack.” So we hadn’t actually started making it yet, and I showed this to new employees when they start at Slack, and it’s in our archive. Because the amazing thing is, it had the pricing that we went live with, it’s exactly the product vision, it’s exactly the way we marketed it.

Everything was preset, and we didn’t have to change it—because, I think, we had three-and-a-half years of practice, and finding product-market fit for our market-of-one team.

 

HOFFMAN: The market-of-one team that Stewart is referring to is the team that was making Glitch, the team that had made the communications tool that would turn into Slack. They’d spent three-and-a-half years creating a hit product, and they didn’t even realize it. But once he had the pitch deck in place, it didn’t take long for him to take action. They were building Slack.

 

BUTTERFIELD: I think all that stuff actually took place in 72 hours, and within a week, we were definitely on this plan.

 

Well, I think that we knew that we had something, just because of the value we got out of it. And the other thing I talk about is going to an Andreessen Horowitz partner meeting, at the end of 2012—I hadn’t been in a while. They led our series B, and we were going to tell them, “Here’s what we’re doing.”

 

And part of the pitch was, “We think that one day, in the fullness of time, if we’re incredibly successful, we will get to $100 million dollars in revenue on this business, and that it could thereby be a billion-dollar business.”

 

And there was a lot of “Pfft, good luck”—not a huge degree of enthusiasm. But, you know, that was our whole ambition, and we exceeded that last year, and we doubled it again this year. We really underestimated how big of a deal it could be.

 

HOFFMAN: How do you go from “Pffft, good luck” to a $5 billion valuation? Pivots and persuasion.

 

BUTTERFIELD: We also underestimated how hard it would be to get people to switch, because no one thought that they needed anything.

 

We weren’t competing with anything else that was out there, other than the use of email, which is pretty much impossible to compete with, and whatever it was that people did—whether that was using a bug tracker, or a wiki, or Google Hangouts, or Skype. No one thought that they needed Slack.

 

HOFFMAN: Its genius is, “You’re a madman, but you’re right.” And this is part of the creating a new market.

 

BUTTERFIELD: We went out and started trying to convince other people to use it, which was way harder than we thought it was going to be.

 

Social proof was a thing that we needed. So we needed to be able to point to teams other than our own, because of course we used our own product. So we had some friends who worked at Rdio—the music-streaming service, now defunct. And I was on the board of a company called Cozy, which is a landlord-tenant app, and I had a friend who had a company called Wantful, which is a gifting service. And we just went to their offices over and over again, and tried to figure out ways to explain why we thought it was valuable.

 

They were friends, first of all, and they were very tech-forward, they were enthusiasts, and they would want to try new products.

 

But what actually made it hard was, you have to get a whole group of people to change all at once; it wasn’t just one person. The benefit of that, though, once you’re able to break through, is once the group starts doing it, what was difficult about it becomes its benefit. It’s very hard to get the group to change to anything else once they switch. So Slack was a really interesting product to scale in the early days, because it had very, very little network effect between companies, but a huge binary, 100% or just nothing, inside of companies.

 

HOFFMAN: So there were companies to cajole, influencers to influence, hearts and minds to win. The overarching driver of Stewart’s success? The ability to tell a great story—one that inspires people to go on the journey with you, as you make some seriously bold moves. It’s one of his entrepreneurial superpowers.

 

BUTTERFIELD: If there was one piece of advice I wish I could phone back and give to myself, was just concentrate on that storytelling part, on the convincing people. Because if you can’t do that, it doesn’t matter how good the product is, it doesn’t matter how good the idea was for the market, or what happens in the external factors, if you don’t have the people believing.

 

HOFFMAN: When Stewart knows that it’s time to slash and burn the business, the only way to lead your team with you on the journey is to turn it into a great story. Find those moments where everything changed, whether it’s throwing up in the hotel bathroom, or having a mountaintop moment you have to bring to your team. Figure out what your story is, and it will go a long way to bring your co-founders, employees, and investors along with you.

 

And you’d be surprised who might come back to help. Remember Tim Lefler, the engineer that Stewart had to lay off? He was naturally a bit reluctant to take another job offer from Stewart.

 

LEFLER: I got an email from Stewart—and getting an email from somebody who previously laid you off, talking about work, you’re a little cagey on that. But there was no reason I wasn’t going to meet Stewart for coffee. He gave me the rundown of what was going on, and just said, “Hey, we’d like to have you back and working on this product with us.” I had decided I was going to stay out of the startup world again, and this was another risk.

 

I talked it over with my wife. My wife is quite the opposite of me when it comes to risk. She said, “If you were to leave where you’re working and go work for Slack, and Slack fails, how sad are you going to be that you need to find another job?” And I’m like, “Yeah, it would suck, trying to find another job. And the job I might get might be worse than what I’m doing now.” But then she said, “Well, what if you stay where you are, and Slack becomes a huge hit?” And I basically said, “Well, I would feel really bad about that.” And so that was the clinch decision for us, and we ended up moving.

 

HOFFMAN: Yes, Tim and his family moved again. Fortunately, it was the right move this time around, and when he arrived at the office, he was surprised to find that he wasn’t the only battle-scarred employee to rejoin the team.

 

LEFLER: I came back with a bunch of people that had also been cut, and were coming back for the same reason. It felt like a bit of a reunion. It was a nice feeling, kind of a familiar feeling.

 

HOFFMAN: So what seemed like the end of the road for Stewart and his team, was, in truth, only the beginning. Bear that in mind if you ever have to slash and burn your business.

 

I’m Reid Hoffman. Thanks for listening.

FOLLOW US