The co-founder of LinkedIn, partner at blue-chip venture capital fund, Greylock Partners, Microsoft board member--and Forbes 400 billionaire--talks about his podcast, "Masters of Scale," and why hypergrowth tech companies need to play by different rules.
Q: You’re a busy guy--venture capital partner, serving on such boards as Microsoft’s. Why do a podcast?
Hoffman: I taught a class on what I call “blitz scaling” at Stanford a couple of years ago. There are so many amazing technology companies being built, and I wanted their founders to share how they rapidly scale their companies. How they manage their people, customers and product development, all while making the incredibly rapid changes they need to make.
Q: Why would any founder share his or her successful secrets?
HOFFMAN: Part of what we have here in Silicon Valley is this network of people talking to each other about the key lessons to learn. We believe in bringing entrepreneurial knowledge to the world.
Q: You like to invest in companies with network effects. Without those network effects, is there any reason to grow your business like crazy?
Hoffman: Most Silicon Valley kinds of businesses have network effects. There’s a real first-mover advantage. You have to move fast. If you want to transform the world, you have to have scale. Funny thing is, you have to embrace inefficiencies to get speed and scale, and that runs counter to traditional business thinking.
Q: Explain.
Hoffman: Traditional thinking is about efficiency: Get your unit margins right; hire and adjust your organization for very little chaos. You learn this at Harvard and Stanford business schools, or at GE and Disney. With blitz scaling the only things that matters are rapid engagement and growing the customer base. You can refactor efficiency later.
Q: How do you keep customers happy when you’re growing like crazy?
Hoffman: Traditional thinking says that with your customer service you should be really high-touch; your customers should always love you. But blitz-scaled companies in Silicon Valley and China do things in a different way. They say, “Well, we’ll just have e-mail customer service. We won’t have any phone customer service or high-touch stuff, because we can actually scale e-mail customer service much faster.”
Q: Blitz-scaling works for digital companies. Will it work in physical industries like manufacturing and transportation or in regulated industries like banking and pharma?
Hoffman: I think it will. And, by the way, going into a regulated industry isn’t new. I mean, I did it at PayPal, which challenged banking. Transportation with Uber is regulated. Lodging with Airbnb has regulations. So, yes, I think we’ll see more blitz scaling applied to traditional, slower industries, whether they’re in construction, education or medicine.
Q: Talk about the culture of blitz scaling. You’ve interviewed Reed Hastings, Sheryl Sandberg, Eric Schmidt, Selina Tobaccowala. What do they say about culture?
Hoffman: They all say successful cultures operate on first principles. Speed means you have to change and adapt tactics constantly. You can’t do that if you’re rules-based.
Q: There’s a lot of criticism today that tech companies reward jerks.
Hoffman: No question. The important thing is to be deliberate and explicit about your culture. The CEO and management team must live the culture, so that everyone else lives it. In successful tech-company cultures, every employee is a culture standard bearer.
Q: Examples?
Hoffman: Netflix. It shared the problem of every Silicon Valley company: High-talent people leave because they don’t fit the culture. So Netflix said, “We need to do better at screening who fits. We’ll do this in the interviewing process.” But that didn’t work. Then Netflix said, “Well, let’s just broadcast what our culture is. Then the right people will come here, and the others will know they don’t want to work here.” That’s worked well for Netflix.
Q: Your favorite podcasts so far?
Hoffman: Mark Zuckerberg. I started the conversation by saying, “Congrats, Mark. Facebook finally grew up. You changed Facebook’s cultural principle from ‘move fast and break things’ to ‘move fast with stable infrastructure.’ Why did you change?” And he looked at me and said, “Nothing changed.” I’m like, “Well, your words changed. ‘Move fast and break things’ isn’t the same as ‘Move fast with stable infrastructure.’” Zuck said, “No, no, the point is speed. When you’re large, and you start breaking the infrastructure, that’s actually going to make you slower. You don’t want to always be rebuilding infrastructure, because that will actually ultimately be slower.”
“Wow,” I thought. “Zuck is smarter than I’d thought.”
Q: I want to switch subjects. You’ve recently written some provocative papers on the future. In one you said you believe that people will be fully happy to give up their right to drive or, at least, to give up driving in autonomous zones. I think the real world will react differently. If gun owners don’t want to give up guns, I can’t imagine people giving up the right to drive.
Hoffman: I think there will be a bunch of people who will say, “I like driving. It gives me a sense of power, empowerment through this skill I know how to do.” But I do think that most people, once they see the virtues of autonomous driving, will say, “Oh, maybe I don’t have to own a car. Maybe I don’t have to drive around looking for a parking spot.” I mean, do people miss their CDs?
Give people a choice, and car owners will suddenly see the hassle of all the little things that go into buying, maintaining, insuring and dealing with the things that break down, and all the rest that goes with having a car. So I have absolute conviction that the vast majority of the human race will get there, once the technology is ready. It may take a bit of time to convince the old guard and put things in place. But the benefits are just stunning: 90% reduction in fatalities, no more gridlock.
Q: You wrote that drinking and driving will no longer be a problem but an opportunity. Explain.
Hoffman: Exactly. Because you’re not driving, you can have a nice meal and a glass of wine during your commute.
Q: One more question on autonomous driving. Litigation. You can imagine trial lawyers lusting for the opportunity to sue Uber, Google or any rich tech company over any accidents caused by driverless cars.
Hoffman: I don’t disagree. There will be a bunch of things, in addition to getting the technology worked out, that will be important. That gives us more time to work through other societal issues, such as workforce transitions.
Q: Such as the fear that driverless trucks will end employment for a lot of truck drivers?
Hoffman: Well, most truck carriers are already having difficulty hiring enough truckers. That’s why Greylock is leading the investment in Convoy. But sure, there will be a lot of societal issues to work out.
Q: You also wrote a paper about the role of AI in management. Talk about that.
Hoffman: First, there’s a big difference between specialized AI and general AI. Specialized AI is here already. Part of what my article was saying is that we can already begin to work in much better ways by deploying specialized AI. Example: “Are we getting information to the right people? Are we inputting the right information? Are we making the right decisions? When decisions are made, are they propagated in the right way?”
Another pretty mundane example is meetings. How do you set up meetings, decide who’s in the meeting and how it’s to be run, how do you take notes during the meeting, how do you prepare for the meeting, how do you set an agenda? You know, all of these things are part of professional business practice, and meetings are such an important part of any business.
Specialized AI can handle the challenge of making meetings more productive. It can say, “Okay, you’re the note taker. You make sure the notes are distributed to all the right people.” It can record what goes on in a meeting in terms of action items. For example, it can record that Reid is to follow up with Jeff. And, you know, that’s like a reminder: “Hey, Reid, you said that you’d follow up with Jeff on that,” which is easily emailed to Reid’s--and every other attendee’s--Outlook.
Small improvements add up. How much better do your companies work when you have better decisions and followup? Even when it’s just 2% or 3% better information coordination, sense of participation or line of sight?
That’s easily done today. For the future, how can both specialized and generalized AI create symbiosis? How can we work better together? How can we create a higher-functioning symbiotic organization?” How can we integrate higher-functioning teams and organizations into the human ecosystem with positive benefits?
Q: Speaking of higher-functioning teams and organizations, you sit on Microsoft’s board. What do you think of Satya Nadella and his performance in reshaping Microsoft?
Hoffman: I think Satya is perfect for the business. High IQ, high intellectual curiosity, high drive. He’s one of a relatively few top CEOs who, when I sit down with them, ask a bunch of questions. I think curiosity is a key to succeeding in the modern age.
