Telecommuting and Yahoo's Desperate Need for Innovation
Management expert John Sullivan says Yahoo is right to end telecommuting
Steven Cherry: Hi, this is Steven Cherry for IEEE Spectrum’s “Techwise Conversations.”
Yahoo’s CEO, Marissa Mayer, and its executive vice president of HR, Jacqueline Reses, started a firestorm a few weeks ago with a memo that called all its work-at-home personnel back into the office, beginning in June. By some reports, that’s only a few hundred employees, but it seems others, who work at home regularly for part of the week, or who less regularly but more than occasionally work at home, would no longer be allowed to work remotely as well.
On the one side of the argument are voices like that of business tycoon Richard Branson, no stranger to managing tens of thousands of employees, who quickly blogged on the matter and tweeted, “Perplexed by Yahoo stopping remote working. Give people the freedom of where to work, and they will excel,” and Matt Mullenweg, founder of WordPress, who commented that he “is 100 percent committed to being distributed—130 of our 150 people are outside of San Francisco.”
Valley culture isn’t entirely in favor of telecommuting, however. It’s the exception at Google, which is not only a model of high-tech success in general, but where its former employee No. 20, Marissa Mayer, had spent most of her professional career.
There’s no way I, or any tech journalist, or even any tycoon or start-up founder is likely to know what’s best for Yahoo better than its CEO does. But I thought we could step back from the firestorm and ask: What do we really know about telecommuting, a practice whose efficacy is a moving target, given rapid advances in computing, broadband, mobile communications, and the technologies of collaboration?
So my guest today is John Sullivan. He’s a professor of management at San Francisco State University and head of his own company, DJS Consulting. He joins us by phone.
John, welcome to the podcast.
John Sullivan: Thank you, Steven.
Steven Cherry: Let’s start with a few basic assertions, pro and con, that are often made. No. 1: Employee interaction encourages innovation, while working from home boosts productivity. True or false?
John Sullivan: True.
Steven Cherry: In an op-ed piece in The New York Times, a Silicon Valley entrepreneur argued this: “But it is also the case that some of the most creative insights come only when you give the human brain unstructured time to think. Opportunities for such freewheeling thought rarely present themselves amid the hustle and bustle of daily office life.”
John Sullivan: Well, the key point is you can’t mix, you know, words. Creativity is not innovation. Creativity is idea generation. And, you know, you might create a lot of ideas smoking pot or sitting in a room, but innovation is implemented ideas. Innovation is ideas that come to marketplace. So you’re not comparing the same thing. If you look at telecommute workers, they’re happy, they’re engaged, they’re productive, but none of those are even related to innovation.
Steven Cherry: I mentioned that the initial target of the Yahoo edict was its customer service personnel. How much innovation do you think Yahoo gets from that group? Why not have one rule for them, say, and another for product teams?
John Sullivan: Well, you could have a train where one part of the train’s moving at 5 miles an hour and the rest at 50, and you’d have a crash. So the answer is no one’s exempt from innovation. You have to remember, Yahoo is fighting Google, which is a machine—the No. 3 most valuable company in the world, a serial innovator unlike any other except, perhaps, Apple, both of which are within, you know, bicycle distance of Yahoo. So they’re getting their ass kicked, and they have been for many years.
And so the answer is, the only way to beat Apple is to out-innovate Apple. The only way to beat Google is to out-innovate Google. And since Marissa comes from Google, as you mentioned, she knows they have tons of data to show that there’s a formula for innovation. There’s a way to improve innovation. There’s a way to become a serial innovator. And that is totally different than being productive. Accountants are productive, but they’re not innovators. And they will probably actually hurt your company.
So Larry Page, the head of Google, actually has a phrase: He expects you to improve at 1000 percent. In other words, improve everything you do by 10 times. And he actually says if you focus on continuous improvement—or efficiency or productivity, in my terms—if you focus on continuous improvement, you are guaranteed never to be wildly successful because you’ll be so focused on improving by, you know, these small percentages, you’ll never see the big picture. So a company will, you know, miss the opportunity that eBay or Facebook or Twitter came along with.
So you have to be careful. The profit return from innovation is literally 10 times higher than the profit return from efficiency. Accountants make you very little money; innovators make you lots of money. For example, the average employee at Apple generates US $2.2 million every year. The average employee at Yahoo generates $350 000. Six and a half times. So what Marissa’s doing is saying, I need this kind of output from my people. I need this kind of profitability, and it only comes from serial innovation. We’re not a serial innovation firm. And every part of it from customer service to HR to janitors have to be part of it.
And the Google data shows, incidentally, that it’s not interaction between engineers; all people at the same function work together, just a common practice. It’s interaction between engineers and finance, engineers and design, engineers and production or marketing. Those are where the interactions—they’re called “serendipitous interactions” or “casual interactions”—that’s where the collaboration and the innovation comes from.
Steven Cherry: Sure. I’ll ask the whole thing. What about the Richard Branson credo: “Give people the freedom of where to work, and they will excel.” Let’s call this assertion No. 2: Employees who are empowered to choose where and when to work will work more happily and better. True or false?
John Sullivan: Well, those are what we call words. So words are interesting, and they make good poetry, but you need data. So he has no data. He’s not a data-driven guy. Google has data. So they have a whole function called the People Analytics group that does nothing but analyze how to get the most out of workers, how to increase innovation. So whenever anyone has an opinion, I mean it’s great to have an opinion, but if you’re an engineer, you understand that at Google or Facebook, decisions are made based on data.
So the data is crystal clear, whether it’s a university study from Harvard Medical School or from internally at Google, is that the profit return from being innovative is so high that it’s worth any cost—free food, free bus shuttles, free you name it—compared to the actual loss when you focus on continuous innovation.
So he’s wrong, and he has no data to back up his statements. He’s equivalent to Donald Trump. He has an opinion. Good for him.
Steven Cherry: I guess that’s also going to take care of assertion No. 3, which is companies that encourage working from home are saving a fortune in real estate and related costs. True or false?
John Sullivan: No, they’re losing a fortune. The real estate costs are a real savings, but the loss of innovation is the real cost. So one more time, the most valuable firm—the No. 2 most valuable firm in the world is Apple. The No. 3 is Google. Neither one allow you to telecommute. The highest IPO in history is Facebook. They do not allow you to telecommute. Maybe there’s something to that.
Steven Cherry: So if what we really care about is things like innovation, why don’t we measure it directly and give managers the freedom and the responsibility to get it whenever and with whatever policies that work?
John Sullivan: Well, because in order to be a manager, you sometimes just have to be there longer. And maybe they’re even an engineer, but they’re—they didn’t really take a class in management. So unfortunately, in the high-[tech] field, there aren’t many good managers. So if you look at what Larry Page has done at Google, they’ve turned management into science. And there was even a time at Google where they tried to discourage having managers. But because they’re a data-driven math camp, they did an analysis data program called Project Oxygen. They analyzed hundreds of—or thousands, actually—of examples and said, do you need managers? If so, what makes a great manager, and can you turn a bad manager into a good one?
And again, using a data-driven approach, they found that there are eight characteristics of a great manager, you can turn bad managers into great leaders, and that the dollar impact of a manager is the highest of all of the factors. So, you know, it’s a data-based decision. This is a math camp. You make decisions based on what works, and you don’t have opinions, you know.
Steven Cherry: There were some stories that Marissa Mayer was relying on VPN logs that showed her that people working remotely weren’t logged in very much. That seems pretty unlikely, because VPN logs probably aren’t a great—uh…What other data do you think she was bringing over from her Google experience that led her to think that?
John Sullivan: Well, all of it. I mean she’s even brought Google people with her. She’s not dumb. And so they have diversity studies that show how to get more women engineers, how to retain them, how to get them promoted, how to get them more productive. They have a really intriguing retention study, which can predict when an engineer’s going to quit six months in advance with a, you know, 80-something percent accuracy rate. They have an algorithm for hiring. They know precisely what kind of person will succeed, you know, if they’re hired. Google is an incredible place. So she just borrowed it, and I’d recommend everyone do that.
So, you know, at Google they measure the length of the cafeteria line. They measure the length of the table in the cafeteria. Why? Well, if the line’s too short, people will just get their food and they won’t talk to anyone. If the line’s too long, people will get frustrated and leave the line. So the goal is three to four minutes in the line, have you collaborate with different people. If you sit at round tables, it turns out you talk to only two people, but if you walk in and there’s a long table like in high school, you will bounce off or interrupt or talk to eight people. And so they have designed the building, the tables, the coffee line, everything, in order to increase collaboration.
So there’s a formula, which not everyone knows but, you know, it’s not a top secret, is that at Google, innovation comes from four—three factors. First, rapid learning, or what they call “discovery.” Second, collaboration, which occurs when you bump into people. And third is fun, which when you’re having fun and you bump into employees, you learn and discover. So they have a formula for it. They—you—you have a lot fun there. And by the way, you have the same fun at Facebook when they have a free ice cream store, free bakery, free barbecue. So it’s all scientific.
So when I hear these uneducated people with opinions, it’s like, no, you’ve never been to Google, and Facebook copied exactly what Google did, and Marissa’s copying exactly what Google did. So it’s the No. 3 most valuable company in the world, and its IPO was, what, nine years ago? It’s a phenomenon. It’s unbelievable. People could learn a lot from using the data.
Steven Cherry: One thing we haven’t talked about is a possible middle ground. The New York Times op-ed piece that I mentioned before described an in-office policy of 12:00 noon to five o’clock, three afternoons per week. Leaving those details aside, is it possible to get the best of both worlds?
John Sullivan: Sure. You could have half the innovation if you let people work at home half the time. No, that’s silly. So op-ed means opinion; it’s not data-based. So if you can find a number in that op-ed, let me know. But the answer is no; this is science. So if your audience is engineers, you ought to be angry about these people in HR or social work or wherever the hell they come from with opinions. This is scientifically based. There’s no accident. She’s doing this to double, triple, and quadruple profit. You can’t innovate half the day. And if you could, why, when you could innovate the whole day?
Steven Cherry: Well, you’ve provided a lot of ammunition, I think, for the anti-telecommuting crowed and a lot of ammunition for my personal campaign to get free lunches here at Spectrum.
John Sullivan: Well, how’s that? When you bring people in, you have to pay real estate, as you mentioned, but the food is expensive. They ride the shuttle. It’s incredibly expensive to bring people in. She would have to be an idiot to spend all that extra money if there wasn’t a huge return. And that’s what people don’t realize. If you can innovate, serial innovation like Apple, there is so much money that the food becomes, you know, chump change. It doesn’t matter.
So remember, this is a place—Google has a bowling alley, free massage, all this stuff that every other company in the world thinks silly because they don’t have the data. If you had the data, you’d say, hey, let’s have two bowling alleys.
Once you understand the importance of interaction, the only problem with interaction is that you have Google employees sitting—so they do open office, I’m sure you’ve seen their standing desks, all that stuff. So the only—you interact with people and you run into someone and say, “Steven, whatcha doing?” “Oh, I just won the Nobel Prize.” Well, now I feel stupid slow, and I say, “Well, how’d you do that?” And you can help me with my problem. So that’s where the value of innovation comes. But what you haven’t thought of—what if the people you ran into didn’t work for the same company?
So the next step of what Marissa’s doing—Google, of course, being the leader, already does it—is what’s called “co-work,” where you will literally not telecommute, but you will be sitting close to people from other companies, from faster, mostly start-ups, more agile companies, so that your learning speed, for discovery, your interactions, collaborations, will be literally twice as powerful. Your solutions outside the box will be from a much bigger box.
So the next wave that most people haven’t heard of or haven’t thought of is co-working, where you get the same amount of interaction, but you also get it with people who don’t work for the same company, and therefore you increase innovation even faster.
Steven Cherry: That actually sounds a lot like MIT’s Building 20, which was famous for the way unplanned encounters between seemingly unrelated academic departments there led to some world-class discoveries in high-speed photography, microwaves, Chomskyan linguistics, and I believe the first video games, so…
John Sullivan: And so, you know, we’ve had Bell Labs. We’ve had PARC. We’ve had all sorts of famous laboratories, but we’ve never had people from different companies innovating and doing research side by side. And that’s the next iteration. You’ll be doing a podcast on that a year or two from now.
Steven Cherry: Thanks for your time today, John.
John Sullivan: All right, Steven.
NOTE: Transcripts are created for the convenience of our readers and listeners and may not perfectly match their associated interviews and narratives. The authoritative record of IEEE Spectrum’s audio programming is the audio version.