GE ranked ninth on Bloomberg BusinessWeek’s 2013 list of the world’s most innovative companies. Not bad for a business that rated number eight in the same year’s Fortune 500 list of the world’s largest enterprises — and positively astounding for a company that was founded in 1892 and now boasts over 300,000 employees and nearly $150 billion in revenue (fiscal 2012). Much of the credit goes to their CEO Jeff Immelt and GE leaders such as Stephen Liguori. As GE’s executive director of global innovation, Liguori and a dedicated team are pioneering the use of lean startup techniques in the arena of industrial hardware. His FastWorks program is driving lean startup practices throughout the company to stimulate breakthrough products and open new markets. Liguori and his colleagues are solving the most intractable challenges of enterprise innovation through bold leadership. He spoke with us about the path he’s taking to get there.
How are your innovation efforts structured? Do you have a special innovation division?
We don’t have a special division. There are two halves to the innovation equation at GE. The first is technical innovation, inventing new machines. We have six Global Research centers worldwide, including a new one with 700 software engineers in Silicon Valley, who work with every GE division. The other half is commercial innovation. Beth Comstock, our CMO, heads a small hit squad doing that. I’m the executive director for global innovation. Sue Segal is president of GE Ventures, and there’s a licensing team that looks for ways to take advantage of the thousands of patents we have.
How do you keep innovation efforts free of interdepartmental politics, budgeting cycles, and other corporate roadblocks?
It’s the heart of the issue. When you combine the bureaucracy that builds up in a large organization with the highly technical nature of the things we make — jet engines, power turbines, CAT scan machines — you could say it’s daunting. Our solution is FastWorks, a program built on lean startup principles. If you don’t recognize that the culture inadvertently gets in the way, you’ll lose. You’ve got to go top and bottom. At the top, you have to get not only sponsorship and also buy-in and understanding. We’re putting executives through training in lean startup principles, telling them, “You have to know how to do it because if you’re not going to change your behavior” — that’s literally what we’re talking about — “we’ll get crushed by competition.” The flip side is giving the teams tools and training. They say, “I want to be an entrepreneur, but I get choked to death by the functions. The finance guys say prove it, the legal guys say it’s too risky, the compliance guy says the regulatory people will have problems with it.” Now we need all those functions, but we have to quickly evolve.
How would you answer them?
One of the ways you get the buy-in with the bureaucracy and the culture is by telling everyone, “We’re not betting the ranch.” We’re not going to build a factory to roll out thousands or millions of new, innovative machines. We might make a million refrigerators a year, but we’re making only 60 of the conceptually disruptive refrigerator we’ve been working on. It’s not the old, “Give it to engineering and come back three years later” — and, in our case, $30 million later. It’s like, “How about we give you $30,000 and 30 days and you come back with a prototype.” Then the engineers say, “Do you know how much money it will cost me to set up to make a prototype? My yield will go down, my waste will go up, my metrics will go in the toilet and I’ll get really dinged at my annual review.” It’s not that the incentives are set up wrong; they’re set up great for running the mother ship. This is not the mother ship. We’re telling the teams that the “minimum” in a minimum viable product is not just in terms of the features; it’s also the smallest number of customers who need to use it to get real learning. You might make five prototypes of a locomotive and give one to each of the big north American railroads — just one, not 50, not 500, – and that contains the risk to the system. We’re doing it to discover a need in the marketplace, not to scale yet, not to make money yet. That opens the window to get people to listen to these radically different ideas. “We don’t know if it’s a big idea yet, so we’re not looking to blow up the world. We just are looking to do a small test, and oh by the way, do it faster!
How do you think about staffing for innovation?
There are only five of us on the dedicated HQ FastWorks team. A senior group of execs asked us just the other day, “Do you need more people to do this?” We said, “No.” It’s probably the first time at a GE meeting where someone was offered more resources and said no! There are two reasons why we did that. First, there’s not a startup in the world where they don’t talk about the scarcity of resources. We’ve learned that it’s not about how much resources you have; it’s about having the right focus. Second, if this becomes a mandate from headquarters, it will fail. We’re spreading the word, GE division by GE division. There’s a team from GE’s energy business, they make unbelievably huge, complicated turbines that go in power plants, plus every piece of what’s called transmission and distribution. It’s just mind-bogglingly complicated. They want to try FastWorks on three new product ventures to get into a radically new space. Here’s the punch line: They’ll staff it and fund it. Our HQ team will bring in an entrepreneur who can coach them in how to be startup-like in GE. We want buy-in by the teams: “You put your skin in the game. You put your people and money in to fund it. We’ll provide the training and coaching and tools to help you do it.”
Would it be productive to structure compensation like a typical startup, in which founders give up as much as half their salary in return for equity in the projects they develop?
The compensation plans today do not include equity, but we’d love to get into that. I did a Google hangout with a couple of folks on crowdsourcing. We literally crowsourced some jet engine parts, believe it or not. The guy who won was from Indonesia. Second place was from Hungary, and third place was from Poland. It’s amazing the smart people you can find around the globe. The question came up, “How much are you paying this person?” We payed the winner $7,000. Someone said, “Couldn’t GE potentially make millions of dollars from that?” Yes, we could, even though that’s years down the road. So we’re trying to figure out the right incentive system for internal startups as well as for crowdsourcing. Here are the counterpoints: When you’re a startup within GE, you have resources and career stability, but if we don’t provide the right rewards, they’ll leave. So we’re aware that we need to let employees put some compensation at risk so they can have a much bigger payout. We have discussed letting employees put compensation at risk in exchange for upside. The very people who say, “I need more innovators and risk takers” are worried about screwing up the compensation system. Will we go all the way to equity or spinning off joint ventures? I can’t answer that question today, but I can guarantee that GE will be experimenting in the next year or so.
Is there a vision or a thesis that directs or places boundaries around innovation efforts?
We probably have two theses. One is to move from building business models on equipment to building them on holistic systems and solutions. How about, instead of providing airlines with jet engines, we provide power by the hour. We say to a power plant, “We’ll tie in wind farms and solar farms, and the more power we give you, the more you’ll pay us because it lowers your operating costs.” That kind of business is very different for a company that grew up building and selling physical pieces of machinery. The other is that the world is becoming increasingly connected and increasingly kinetic — from a speed standpoint — and we’ll either innovate faster or be disrupted. It’s an opportunity to tap our domain knowledge and heritage around technology and do even smarter things with it. We’ve got to move with market speed or market intensity.
What role does the lean startup method play in FastWorks?
The lean startup method is absolutely one of the key pieces. Jeff Immelt’s annual letter to investors for 2013, he says that two of the best books I’ve read lately are _The Lean Startup_ by Eric Ries and _The Startup Playbook_ by David Kidder. Those are the only two books he mentioned in a five-page letter, which speaks to the fact that the lean startup method is one of the core influences on what we’re trying to do. Every company needs to figure out how it applies to them. We take the principles and morph them to make sense for us.
How do you use metrics? Do you use innovation accounting to track the progress of projects before they earn substantial revenue?
We’re beginning to. Right now, we’re actively working with a dozen or so projects where the goal is to identify metrics and develop innovation accounting techniques.
How do you know when one of your projects has reached product/market fit?
We know we’ve reached product/market fit when the market leaders, the early adopters, say, “I’m in.” That’s not as subjective as it might sound. We’re encouraging people to come up with 10x better solutions: Don’t just be marginally better than the competition, make it 10x better. If you’re 10x better, the customer will say, “I’ve got to have that.” Then we know we’ve got something. We’ve got software to lets hospitals run better; think of it as air traffic control for hospitals. If you put that software in a hospital for a 90-day trial and at the end of 90 days they say, “Don’t take that software out,” we know it has reached product/market fit.
What percentage of your projects is incubated internally, versus acquisitions and investments?
For us, a bolt-on acquisition is $2 billion to $4 billion, and we’ll continue to do those as Jeff Immelt has said. We also do investments with people who can act as partners who get us up the curve quicker. We put $100 million into Pivotal, the VMWare spinoff, and $30 million into Quirky, the consumer appliance startup. The pendulum is swinging toward finding things earlier, inventing them ourselves, and finding partners who synergize with what we’re doing. Look at what we’ve done with our Silicon Valley Software Center. Those 700 software engineers are there fundamentally for organic growth on the industrial Internet. If you go back 10 years ago, GE grew predominately by acquisition. We’re much more balanced now.
What’s the percentage of acquisitions vs. acqui-hires?
We’re in the early stages of looking at acqui-hires. We’ve done a couple of small acquihires to date, and I expect you’ll see more of that.
How can you scale innovation efforts?
GE is unique in its physical scale. It’s partly a matter of training at three levels: executives, coaches, and rank and file. External experts are helping us train a couple of coaches per GE business. We’re up to 80 or 100 coaches who train people on real projects with real GE businesses. Our goal is to scale that as rapidly as we can, but keep the quality going.
Stephen Liguori, executive director, Global Innovation