Let fires burn

CHERYL KELLOND: I’m driving this huge rig that’s like 35 feet long over highways and bumpy roads.

HOFFMAN: That’s Cheryl Kellond, entrepreneur and amateur trucker. Last December, she rented a three-quarter-ton pickup truck and an Airstream trailer – picture a massive chrome Twinkie on wheels. Then she drove some 1,000 miles across Colorado to meet her customers face-to-face.

Cheryl runs a scrappy little healthcare company called Apostrophe, Inc. They have a great tagline: “We make healthcare suck less.” How? By helping employees at small- to medium-sized businesses avoid staggering medical bills. Cheryl couldn’t understand why so many diabetic patients wound up in the emergency room. A simple visit to the doctor would have saved them a fortune.

KELLOND: We’re like, “Why aren’t they going? We just need to get out in the field.” And so it seemed like a great idea – we’re like, “Oh, I know what we’ll do. We’ll rent this Airstream trailer, and we’ll do a lightweight mobile clinic for a week. We’ll invite people in for free biometric testing and quick health checkups, and while they’re there, we’ll talk to them to figure out what’s going on.”

HOFFMAN: Roughly 45 minutes into the road trip, Cheryl had an answer. Patients were hit so hard by unexpected bills, they avoided doctors entirely. She calls the insights from this road trip invaluable. But they came at a steep price.

KELLOND: The trailer disconnected from the truck because the bolt fell out, and then I had to reattach it. And there’s an electronic hydraulic lift, and the wire had frayed. So it starts sparking and smoking, and this small electrical fire breaks out on the hitch, like right next to the propane tank. I’m freaking out. There’s sparks everywhere. I run away from it – the whole thing was ridiculous. There’s sparks and flames shooting out of this electric motor, right next to the propane tank.

HOFFMAN: This is going to sound crazy, but I’ll say it anyway: Cheryl had it easy. I’m not talking about her road trip; that was gruelling. But her decision to put out that fire – it’s a no-brainer. She sees sparks and a propane tank. It’s like, two plus two equals [explosion sound]. If only every decision could be so simple. Short of your office actually catching on fire, you’re almost never going to face a decision like this as an entrepreneur. 

Instead, you’ll face figurative fires – a lot of them, more than you can handle. And you often won’t know which one to stamp out first. I believe if you try to put out every fire, you’ll only burn yourself out. The best entrepreneurs? They let fires burn.

[THEME MUSIC]

HOFFMAN: This is Masters of Scale. I’m Reid Hoffman, co-founder of LinkedIn, an investor at Greylock, and your host. I believe smart entrepreneurs don’t try to fight every fire. They have to let some fires burn – and sometimes very large fires.

When you have a fast-scaling company, there will always be metaphorical fires. Some of them flare up suddenly, demanding your attention. Others build slowly in the background, threatening to spread if they aren’t extinguished. If you spend all of your time fighting fires, you may miss critical opportunities to build your business. You’ll be all reaction, and no action. But if you let fires go on too long, you’ll get burned. Deciding which fires you let burn, and how long you let them burn for, can make the difference between success and failure. 

The fundamental rule of customer service has long been, “The customer is always right.” But in a fast-growing business, you often have to make a choice between serving your existing customers or reaching new customers. And if you’re growing exponentially, it’s a no-brainer. For scale companies, the rule is, “Provide whatever service you can as long as it doesn’t slow us down” – and that may mean no service.

Think I’m exaggerating? Back in PayPal’s early days, we neglected customer complaints, even as users grew exponentially. 

HOFFMAN: And because we had a customer service department of three people, we quickly started going in the hole around 10,000 emails a month. And customers got very frustrated with this. First, a few phone calls started trickling in with customers who had discovered our corporate phone number, and were dialing extensions at random to try to reach somebody. And then very quickly, it became, all of the phones are ringing. And then very quickly after that, probably hours, all of the phones are ringing 24 hours, seven days a week. 

So what did we do? We turned off all the ringers on our desk phones, and started using our cell phones for business. Because as paradoxical as that might sound, where you would say, “We’re supposed to be customer focused; we are listening to our customers.” The problem is we have to treat the future customers, not just the current ones. And if all we did was the current ones, we’d never get to the future customers. 

For example, there was this one guy who felt that he wasn’t getting his payments, and needed the payments. He literally drove from Las Vegas to Palo Alto to come to our front-door office. He asked for customer service, so that meant Laurie went out to go see this relatively tall, bulky gentleman – and came back in and went, “Oh my god, there’s an angry customer here.” And so I went, “Hmm, what do we do?” Because we don’t have security, we don’t have any plans for this, first guy who showed up. And there were others. And so I went, “OK, Jamie Templeton is another six-foot-tall guy. So Jamie and I will go out with Laurie to go talk to this guy.” 

It wasn’t my phone number, but there were people showing up at the front door. And at that point, we put a badge on the front door. 

HOFFMAN: And we let those complaints continue, until one day, we were positioned to solve the problem all at once. We flew out to Omaha, set up a call center. Within two months, a 200-person customer service department was up and running. Problem solved – and I wouldn’t have solved it a moment sooner.

At its core, this is a question of triage, of smart prioritization. One place to start is to think about how you might normally approach your to-do list. It can be a tidy, logical, satisfying process. Brian Chesky, the co-founder and CEO of AirBnB, loves it. 

BRIAN CHESKY: I have this kind of interesting way I do a to-do list. I make a list of long tasks, and I try to be as exhaustive as possible, and then I try to group them. I say, it’s like a game of leverage. What one action can take care of those three? And then I do it again. And I do it again. If you do that enough times, if you have a list of 20 things to do, you end up realizing, “I don’t need to do 20 things. If I do these three big things, the other 20 things will kind of happen as outcomes or outputs of it.” So it’s just like doing fewer, bigger things. 

HOFFMAN: That’s a great way to solve the known problems on a normal day. But what about the great unknowns – the problems that flare up unexpectedly, and on a daily basis? Good luck stack-ranking that mess. I guarantee you, in a fast-growing company at any stage – but especially in the early stages – a list will give you no peace of mind. Yesterday’s wisp of smoke might be today’s five-alarm fire. To really master the art of letting a fire burn, you need nerves, vigilance and practice – lots of practice. 

Selina Tobaccowala is as battle-tested as they come. She’s a serial entrepreneur who’s been disrupting industries ever since she started coding in her dorm room in the 1990s. She’s now the co-founder of a new fitness app called Gixo. Before that, as co-founder of Evite and president of SurveyMonkey, she rode through two punishing and exhilarating waves of demand. 

She can tell you about the blazes behind the scenes of those massive successes. Some fires are big, some are enormous, and one you might call the tech equivalent of Chernobyl. 

So how did Selina master the art of fighting fires? You have to go back to the early days of the dot-com boom, when she was an undergraduate student at Stanford. Selina co-founded Evite with her college friend, Al Lieb. Evite, you may recall, became the dominant platform for online invitations in the late ‘90s. It was also only one of three business ideas that Selina and Al were hatching in their dorm room simultaneously. Those are the breezy, experimental days of a startup, when you can spitball ideas without suffering much blowback from users. Selina and Al might have continued working on three ideas at once, until one day, Selina stumbled on their priorities, quite literally.

SELINA TOBACCOWALA: I am very clumsy, as you know, and I tripped over the cord of our $200-dollar Fry’s computer that was beneath our desk. And I’ll never forget, because then our phone rang and somebody said, “What happened to Evite?” And we immediately plugged it back in, we looked at the database, and we were just surprised about how naturally the product had grown – because there was a built-in viral coefficient into the product that we hadn’t really realized. 

HOFFMAN: Did you know the term “viral coefficient?”

TOBACCOWALA: Absolutely not. [Laughter] At the time, it was more just the fact that, “Oh, well, if people invite other people, other people will see it.” 

HOFFMAN: Now this is the day every founder dreams of – customers complaining about a loss of service. They like you! They really like you!

Before you gush like you’ve won the lottery, you have to consider the flipside of passion. The users who love you are also liable to turn on you. Think about how Selina got a call from that angry customer in an instant. She’s being held accountable, like any other executive. She can’t say, “Look, I’m barely drinking age, and my dorm room is a mess.” As soon as your users can’t live without your service, you’re a mature business – and that’s a heady responsibility. Few startup founders truly grasp what happens after users really like you.

TOBACCOWALA: We did not, at that time – being 22 years old, barely graduating from college – that we were thinking about, “Is this something that could scale? Is this a business that could scale? Is this a product that could scale?”

And so a lot of what we realized is, suddenly we hadn’t actually thought that much about, how do you create hardware redundancy? How do you create a database redundancy? And so we had multiple times where the site would just go down, which was more normal in 1999-2000. But it was still something that we had to learn very, very quickly as entrepreneurs. And just how do you create operational excellence is something that you were learning on the job, versus coming into experience with. 

HOFFMAN: You’d be surprised how many successful startups, the celebrated “overnight successes,” flirt with disaster the following morning. Selina’s experience is hardly unique. Take, for instance, my friend Hadi Partovi, the co-founder of the educational non-profit Code.org. Back in 2007, Hadi and his twin brother Ali, launched iLike, a music discovery app. It went live as an early app on Facebook, back when the social network had only just over 20 million monthly active users. 

HADI PARTOVI: And it just exploded in popularity. 

HOFFMAN: The way Hadi describes it, you’d think 20 million active users turned their heads, all at once, and shouted, “More!”

PARTOVI: We had planned to have two servers running this, to judge interest. And basically within 30 minutes, we realized two servers wasn’t going to be enough. So we immediately doubled it, and then we doubled it again, and then we doubled it again, until our entire server farm of I think 32 servers, we could see that by the end of the weekend we’d run out of servers. 

And it was Saturday of Memorial Day weekend. So it’s kind of crazy. We literally rented a UHaul truck, and called up people just basically asking, “Can we come to your data center and borrow machines? We’ll buy them from you, we’ll borrow them from you, we just need them immediately.” And there was a whole bunch of us basically spending our weekend literally unpacking and racking these machines to get it up and running. 

HOFFMAN: The funny thing about this messy rollout is that it’s a recurring story across Silicon Valley. A product goes viral, the team is caught flat-footed, and they go into a wildly inefficient scrum. You might be wondering, what’s the deal with these entrepreneurs? Hasn’t anyone in Silicon Valley heard of a contingency plan? I assure you we have, and I can only reply,  ”Who has time for a contingency plan?”

This isn’t a quirk of Silicon Valley. It’s a competitive strategy. You don’t just want to be the first-mover into a new market. You want to be the first-scaler. You have to seize every last opportunity for growth – even if it means serving customers on a delusional scale.

LISA CURTIS: So I sometimes talk about founders who have this delusional version of reality. 

HOFFMAN: That’s Lisa Curtis, co-founder and CEO of an energy snack and drink company called Kuli Kuli. Their secret ingredient is a nutrient-rich leaf called moringa; it’s all the range in West Africa. In the U.S.? Not so much – though I’ve been steeping moringa leaves in hot water lately, and I have to say, it’s amazing stuff. Lisa saw the opportunity to scale the moringa market here in the U.S. Whole Foods wanted to partner with her and develop a new energy drink. 

CURTIS: So we said “Yes, let’s do it.” And then Whole Foods said, “And we want to launch it nationwide in January.” And this was June. And I looked at my co-founder – and she does our operations – and she just started shaking her head. And I distinctly remember being on the phone and saying, “Yep, we’ll do it, we’ll get it, we’ll make it happen.”

HOFFMAN: So what happened? Those super-hearty moringa leaves clogged the machines on the factory floor. Lisa’s team scrambled to find a new factory. A snowstorm delayed their first shipment. Tempers flared. Buyers got upset. Then, at last, the big day arrives.

CURTIS: At the same time we were sending out to 435 stores across the country, we also had a batch of samples shipped to us. We did, sort of like a, “Cheers, we did it!” And then we all kind of like made this face of like, “Oh no. This is not what we thought.”

HOFFMAN: The taste was fine, but the texture? Gritty. Cue another few months of panic. They discover another magic ingredient, xanthan gum, to smooth out the texture. And voila, they had a product that was ready to scale – after a massively botched launch. So what’s Lisa’s take on this whole perfectly avoidable mess?

CURTIS: I think it was still the right answer to say “yes.” Obviously, hindsight is 20/20 and if I could have maybe spent more time on the factory floor and been there to really make sure that the product we were getting was the product we had expected, that might have solved it. But I don’t think I would’ve changed saying “yes” and seizing the opportunity. 

HOFFMAN: Kuli Kuli is now available in 3,000 stores, quadrupling its footprint in a year, thanks to that Whole Foods launch. What’s important about this story is that it’s full of unforced errors. And I would argue that those errors are fine, so long as you’re pursuing a massive growth opportunity. 

When you’re moving fast, you can’t say, “Oh, we’re going to study this for six months. We’re going to be perfectly wise and avoid a lot of basic problems.” Because then you’re six months behind – and that’s what’s unwise. Rather than prepare for every possible fire, you have to embrace the wisdom of intelligent triage.

When I asked Selina how she acquired this uncommon wisdom at such a young age, she explained that all fires look the same when you’re young. But one terrible blaze can give you a healthy perspective on the rest. 

HOFFMAN: You were a college student back then. Is the way you fight fires now substantially different than the way you fight fires then? 

TOBACCOWALA: I would say that the biggest difference is probably that I get less panicked about fires when they happen – because you’ve fought so many of them, and you have a much better perspective of, “Where to start?” versus the, “Oh my god, people are depending on me. The site is down and I don’t know what to do here.”

I think that’s probably the biggest change, is understanding that. And also I think not feeling so bad about the fact that you are going to have to throw some resources away in the short term, and as long as you explain that to people as they’re doing the work, they also will understand – versus feeling like you’re always trying to find that perfect solution. 

HOFFMAN: When you’re letting a fire burn, it’s not simply up to you to wage a lonely battle with your own nerves. Your team may start to get the sense that, like the band on the Titanic, you’re fiddling while the ship sinks. You have to show them that yes, you see the problem, and yes, your neglect is deliberate. 

Selina learned to telegraph calm after she left Evite in 2001. She moved to London, where she served as Senior Vice President of Product and Technology to the company TicketMaster. 

TOBACCOWALA: So throughout my career, in most of my reviews, the feedback I would generally get is, “You need to be more patient.” I actually think the thing that taught me the most patience was becoming a mom, realizing that it doesn’t really help to snap at a child – or an employee, obviously. 

But I think that in the sense of the patient side was, “How do you actually calm down a situation” – especially when things are going crazy, or going wrong – and, “How do you get people communicating well?” Because the thing that starts to happen when things are broken, is communication is the first thing to break down. And so that was the thing that I realized was simple. It was back in the day before Slack, it was like, “Get everybody on a conference call, make sure everybody has all the information, and get everyone to be calm and work through the problem.” 

And we had to do that quite a bit at Ticketmaster, when there was massive on-sales. Michael Jackson, U2 was going on sale, we had people from around the world, that if there was a problem, that would get on a call. And there would always be a triage leader, and it was very much, “OK, the first and most important thing is ensuring that there’s fantastic communication.”

HOFFMAN: Now this is exactly the sort of calm, measured management that startups yearn for. And there was one startup in particular in desperate need of Selina’s help. It was a funky little website called SurveyMonkey. 

The founder built an incredibly popular tool for online surveys, and succeeded with an astonishing shortage of everything a company needs to scale.

TOBACCOWALA: The founder of SurveyMonkey is an amazing guy. If you think about the fact that this company started in 1999, in the Midwest, and then he picked up and moved to Portland sight unseen – just because he found the Midwest too cold. But if you look at his ten-year trajectory, it’s amazing. He built the business without a penny of funding. He, I think, invented the freemium model, or at least was one of the first people in the world to use the freemium model. There was essentially two developers, the founder, his brother, and 10 customer service agents – and that was it. 

HOFFMAN: SurveyMonkey’s former CEO, and departed friend Dave Goldberg, was well aware that his team was understaffed and in need of an experienced firefighter like Selina. So in 2009, he started selling her on the founder’s vision. It was an uphill battle.

TOBACCOWALA: You think surveys, like that’s not that interesting. I would say every single person underestimated SurveyMonkey, just because of the way it looked. Every product engineer – and even myself – where you walked in, and the first thing you thought was, “Surveys? Is this that interesting?” Both from my standpoint, and then you looked at the product, and thought, “Well, this looks like it’s from 1999.” And so everybody underestimated both the size of the business, as well as just the true consumer value that the product was actually providing

But when I heard his vision for building out a product where not only were you listening to other people’s voices – and being able to make impact and make change – but also building a platform for a data business. It was truly inspiring. 

Two days before I came to interview, I realized I was pregnant. And I flew back to London, and when I landed, I had an offer in my inbox from Dave from SurveyMonkey. And I wrote him back a note, and I said, you know, “I want to negotiate my comp, and I can’t leave because I’m on a European contract for at least three months – so I have a long wait. And I realize that I am in early stages of pregnancy.” And I wanted to give him that opportunity to say, “Sorry, I can’t make this work waiting three months.” 

And he wrote back within five minutes, and told me that he wanted to make SurveyMonkey a place where you could build a family, where you could create a culture, where we could build a big revenue business. But at the same time, by bringing in people with experience, we could actually also make sure that we could have dinner with our kids.

HOFFMAN: Sounds great, right? Every fast-growing startup sounds great from the outside. Then as you move closer, you start smelling the smoke. 

TOBACCOWALA: There were three of them that were coding – that is amazing for a company doing that much revenue. I mean, the margin of the business was unbelievable. But when you have a technology stack that’s built for three people to code, and then you’re trying to scale an engineering organization, it doesn’t work. 

HOFFMAN: Then, Dave dropped a bombshell. Selina’s about to find out just how many fires were raging in the background. You might be tempted to think SurveyMonkey was wrong to let this go, but you’d be wrong.

TOBACCOWALA: When he was interviewing me, he told me, “There’s no backup.”

HOFFMAN: By no backup, Dave meant the survey data, the bread and butter of the business, could vanish without a trace. 

HOFFMAN: Can you explain what would have happened if a server failure wiped out SurveyMonkey’s data? 

TOBACCOWALA: If the system had gotten corrupted or the data had gone down, the business would have been done – because people come back to the same product partially because you want to see your data and trends over time. And if you lose that, you potentially lose the entire business. 

HOFFMAN: So in other words, Chernobyl. 

TOBACCOWALA: Yes. 

HOFFMAN: Complete catastrophe.

HOFFMAN: Crazy as it sounds, every startup, at some point, will flirt with a Chernobyl. And there’s only one sensible response to a potential nuclear meltdown: think like Rain Man. 

Fans of that movie will remember Dustin Hoffman’s character sitting at the poker table, eyes darting, as he calmly calculates the odds of a winning or a losing hand. I try to maintain that calm rationality when I’m faced with a crisis.

When I’m looking at a fire, I will assign probabilities, I’ll assign vectors of probabilities – is a probability going up or down? Down better, obviously. I’ll say, “OK, what’s the actual damage if it hits?” And then, is it correctable after it hits? 

So if it’s a definite fatal – OK, we have this possibility that the business is just going to be over. Boom. Full stop. Done. And by the way, almost all start ups start in that position. LinkedIn was years before it had a backup database. Now we had backup data that we could reload, so could come back in a day or two. 

But the fail-over system, that was probably six years in. So you look at it and you go, “OK we have this really valuable business. This could kill it. Let’s adjust it.” But, by the way, what’s your probability? Is it 0.1 percent, or .01 percent every day? Well then, actually in fact, you can solve it in three months or six months. It was like, “OK, that doesn’t add up to a lot over time.” If it’s 1 percent per day over 30 days, then all of a sudden you’re 15 percent. You’re gonna be dead in 30 days, 15 percent. “OK, let’s solve it now.” 

So when the SurveyMonkey team sized up the odds of a data loss, they pounced.

TOBACCOWALA: The team fixed it really, really quickly, in terms of, put in at least a copy of the database. We absolutely not only were stressing about the situation, but very quickly trying to rectify it, to ensure that that scenario didn’t happen.

HOFFMAN: And was thinking about the potential Chernobyl – the worst case scenario – invigorating, or distracting, or both? 

TOBACCOWALA: It’s always invigorating, because you have to balance the, “How do we solve something that’s on fire very quickly?” with, “How do we make sure that you’re building out a long-term solution, and not wasting a bunch of cycles?” And that’s always a hard, but very fun, challenging problem. 

HOFFMAN: Chernobyl is invigorating. Now that’s an experienced firefighter talking. Once she had that blaze under control, Selina worked her way down a long list of troubles. SurveyMonkey, like every fast-growing startup, had no marketing plan, no strategy for international users, a rigid billing system that prevented some customers from paying for the service, and a mess of code that made every customization a headache. Selina, in turn, hired engineers, marketers, UI designers, translators – an army of specialists to clear away impediments to growth. 

But what I find fascinating about her story is the fire she left burning in the background.

TOBACCOWALA: Yeah, people definitely called the SurveyMonkey web design ugly. 

HOFFMAN: This wasn’t her personal opinion – everyone told her so. 

TOBACCOWALA: And people kept saying, “Hey, you know, this thing doesn’t look nice. The thing doesn’t look pretty.” 

HOFFMAN: The critiques continued for years.

TOBACCOWALA: But from my perspective, what I was more focused on was the fact that people loved the product, and it was performing. And so I was going to let it be ugly for the sake of being able to actually build a great business. 

HOFFMAN: And so was that one of those cases where you’re like, “OK, I hear you. It’s an important thing to do later – but triage to a tertiary concern or a secondary concern.” 

TOBACCOWALA: I think the standpoint is that people spend a lot of time building beautiful websites, and I think that that’s important in certain types of businesses. When we’re talking about a business when you’re trying to get somebody through a very difficult experience in something more functional, it’s far more important that the product is easy to use – and beautiful and easy to use are not always the same thing. 

HOFFMAN: Although it’s funny, I think Steve Jobs thinks differently – but other than his particular genius, which there’s only a few of. 

TOBACCOWALA: And his products are beautiful and easy to use. Obviously, if you can do that, that’s magic. But I think if you have to prioritize one of the two, I would always put “easy to use” over “beautiful.” 

HOFFMAN: Let’s set aside the “beauty versus ease-of-use” debate. They’re both obvious virtues. Focus, instead, on Selina’s willingness to pit one virtue against the other. Selina can say, “Apple’s products are a thing of beauty. They’re magic. Wouldn’t we all love to have a product as eye-catching as the iPhone? Nonetheless, I won’t fight that battle today, possibly not for years” – there’s magic to that decision as well. We shower so much praise on the entrepreneurs who knock a product out of the park, we forget to celebrate all of their brilliant punts.

Jeff Bezos, CEO of Amazon, is admired for a lot of things – design is not one of them. Look at the Amazon website. It’s not going to win a Webby award anytime soon. I can guarantee you a really good product person looking at that website would go, “I know how to get a much stronger engagement loop on this page.” 

What Bezos knows is that what matters most to people is convenience, price and speed. That’s what customers like, and that’s what Amazon gives them. And Bezos won’t allow anyone to vary from those three goals at all. Even if they say, “But you could do this.” “No, no, no. Convenience, price, speed.”

Jeff and Selina don’t just say, “I’ll let that fire burn.” They commit to their decision, even as the smoke creeps under their door.

HOFFMAN: Can you recall a specific example of a glaringly ugly design flaw?

TOBACCOWALA: So when you finish paying for the product, and you finish checkout, there was this big, massive heart that came to say, “Thank you.” And everybody was just like, “Ah, that’s a terrible image!” But you know, we wanted to retain some of the essence of the product. There were certain pages like that, that people would just sort of shudder at. But that page, as an example, the consumer has just paid you, and so is that the most important page to improve? No. It’s most important to get them back into their experience quickly, to let them finish out their survey.

HOFFMAN: My fellow investor at Greylock, Jerry Chen, is unflappable in the face of complaints. He’s a former vice president of VMware, one of those behind-the-scenes tech companies that allows you to connect and secure applications across devices. Jerry set an unusual measure of success for his team.

JERRY CHEN: The worst thing is when you launch a product, and no one cares or notices. And if you have sales reps complain, or customers file bugs against it, then two things. One, you’re releasing while you’re still embarrassed, a la your rule, and B, people only complain when they care. And so I always set as a metric, how many complaints did we get?

HOFFMAN: As a positive metric.

CHEN: As a positive metric, because people are engaged.

HOFFMAN: You need fortitude to accept these fires as a byproduct of growth, because if you leave one fire burning long enough, it will merge with another fire. And that’s when your strategy of intelligent triage will be sorely tested. SurveyMonkey’s web design, for instance, wasn’t just an eyesore for users, it actually hampered Selina’s ability to recruit talent.

TOBACCOWALA: It’s like, don’t judge a book by its cover, but people were sort of judging this product as being a non-Silicon Valley-based, non-beautiful site, and underestimating both its consumer love and its financials. 

HOFFMAN: Given that it was affecting recruitment, did that put pressure on you to try to address the design, or did you have some other way of trying to overcome the recruitment challenge? 

TOBACCOWALA: Once we got people in the door, we were able to get them excited. But the hard part was, how are you getting people in the door? And so it took more effort, whether it was going to PyCon meetups, whether it was going to PyLadies, whether it was going out there and just talking to people, talking to engineers, talking to product people, using your network to get people to enter. And so that was a bigger challenge for us. And part of that, again, was because people were judging what the business was, what the product was – and until you were able to interact with them and pitch them, you weren’t able to break that barrier. 

HOFFMAN: Selina hints at another benefit of letting fires burn: you want employees who will rush into a burning building alongside you. And there are two ways to spot these employees. You can engage in all sorts of behavioral questioning. You know, “Suppose you had 10 seconds to put out an electrical fire next to a propane tank. How would you handle it?” Or you can just let the smoldering parts of your business speak for themselves – and this is key. Most of us are willing to fight fires; it’s a smaller subset of people who are capable of noting the presence of a roaring blaze that might soon cut off all escape routes, while staying focused on the blaze that’s about to consume them.

And this isn’t simply a matter of hiring battle-tested employees, because even the most experienced firefighters will occasionally miss a five-alarm fire in the making. 

Selina is well on her way to scaling her third startup, Gixo. She launched the app this summer. It’s an ambitious venture to bring live fitness classes to your smartphone. And despite her experience, she readily acknowledges she hasn’t seen it all. She knows her blind spots. And she’s already hiring employees who can help her scan the horizon for smoke signals. 

HOFFMAN: So what would you say, in order to accelerate your younger self’s learning of which fires to let burn or not, what advice would you give your younger self? And one is, not everything is fatal – don’t panic. But what would the other things you’d say are, here are some of the principles I now apply that I would want you to learn faster?

TOBACCOWALA: So I’d say some of the principles: one is, is that it is important to do that learning fast. Putting things out the door fairly quickly, and getting consumer feedback, even to a percentage of the audience – that far helps avoid fires. If you get some feedback from an audience, you start seeing what’s happening to the numbers, and that makes a huge difference in terms of avoiding fires. 

I’d say that the second thing, in terms of avoiding fires, is hiring the right people. You can’t expect that you – especially at 22 or 23, but even now – are going to know all the different things. I mentioned before, design’s not a strength of mine. One of the first hires at Gixo was a great designer. And so it’s balancing your own weaknesses with great people, because you know that they’re going to have the ability to see those fires ahead of you. 

HOFFMAN: I’m Reid Hoffman. Thank you for listening.