Imagine It Forward

by Beth Comstock

  With the early days of FastWorks, our plan was simple: every ninety days, new projects would be judged by a growth board on whether they were meeting their goals, and whether they should be killed, pivoted to a new direction, or given another ninety days to move forward as they were. It was classic VC metered, or milestone-based, funding, in which you fund based on progress. It allows you to kill a project early if you determine it’s not succeeding. You are “de-risking” innovation by spending less money on more ideas earlier, and killing off the ones that aren’t working. It allows you to move forward with confidence when an idea is ready to scale, and invest accordingly. David Kidder proved particularly helpful as a growth board catalyst—helping us operationalize the process, and as a coach for business leaders, giving them room to test methods and make them their own.

  My secondary aim was to remove what I call entitlement funding. It’s hard to get a project funded in a big company like GE—there are many people who want to weigh in on a seemingly limited funding pool. But it’s equally as hard to get it “unfunded.” We needed to drop the barrier on both. Innovation without constraints is no blessing. The goal is to think of funding more like an extension of credit—it’s only good until you get to the next phase.

  With growth boards, we could bankroll more small projects, more “productive failures,” to drive a portfolio view of all of GE’s products in development. Growth boards are funnels. In the start-up world, only one in fifty ideas makes it; the growth board process makes sure you kill the other forty-nine quickly. If not, they become “zombies”—(another Kidderism)—the walking dead.

  Durathon became a zombie. But it didn’t have to. It wobbled on for three more years. In 2014, the Durathon plant was idled due to manufacturing trouble; a few months later, GE reassigned most Durathon workers and stopped production. In November 2015, we announced the closure of the Schenectady plant, after investing nearly $200 million.

  Durathon was a classic example of what can happen without the lean start-up concepts, the no-idea-before-its-time mentality we were infusing with GE Ventures, and the growth boards’ rigor.

  Just as we were changing how GE approached innovation, we had to change our training and indoctrination system to be something more emergent. The old GE way was command and control, top down. But we couldn’t afford to treat our new OS with the same mentality, slapping lean start-up and FastWorks and growth boards in a manual and teaching it only in a classroom. This was hard stuff, and we needed buy-in. People needed to engage in it to embrace it.

  How to Organize a Growth Board

  A formal growth board serves to deliver discipline around key decisions in funding, testing, allocating resources for projects, and new businesses. I believe growth boards are fundamental to driving innovation in an organization.

  Here is the framework for organizing a growth board:

  Convene a cross-functional group of decision-makers with enough different perspectives to add value; the smaller the group, the better. All decisions are made here; there are no backroom deals.

  Leverage your core set of questions as projects move through the various stages.

  Ideas/projects are greenlighted only as they move through each stage, meaning they pass the test for should versus can; gaining traction is moving from one to many customers; scale means that the products work as described and the commercial model is sound and growing.

  Beware that growth boards don’t become overly bureaucratic or that you start adding growth boards beyond where they are useful. In venture capital, Mondays are set aside to convene partners, make funding decisions, and send pitches back for more work or decline them completely. VCs are pitched thousands of ideas a year; only a few dozen are funded.

  I use this approach for funding any project with my teams, not just for new businesses. For example, when digitizing workflows or our sales teams across GE, I convened a group of commercial leaders; we agreed to jointly pool our resources—budget, people, capabilities—in a way that would scale the best ideas fastest. Everyone got smarter in the process, and since it was all GE’s money to begin with, we quickly shed the limitations of fighting over budgets and control. I wish I could have done this earlier in my career, but we needed some examples of success with growth boards at GE for people to feel confident it would work.

  GROWTH BOARD PROCESS

  • Discipline around test & learn

  • Capital allocation

  • Right people for right stage of development

  Certainly, we made sure the top leadership teams in each business had been immersed in the concepts and understood the critical questions and decision-making processes. We then deputized coaches within each business unit—sometimes they were people from the unit who had shown passion and skill at the new OS. Sometimes they were outside sparks such as Eric, David, and Aaron Dignan and increasingly other serial founders—entrepreneurs in residence—that Sue and the Ventures team were able to attract and direct.

  Our central teams created a series of communications and best practice sharing in this community of coaches. Because it was people inside the units teaching one another, it gave those in the process a sense of agency and a sense of local identity: “This is ours.”

  The coaches started small, working on the first promising FastWorks projects with what former President George W. Bush would call the Coalition of the Willing. With every success, we’d further congeal this coalition, and the number of willing would grow. Among the early successes:

  Digital Wind Farm, an effort to maximize wind output via software and no additional capital expenditures. In less than four months, the team created a new business model that was validated by a customer and delivered 20 percent more renewable energy per wind farm and a turnkey solution with a revenue opportunity of $50 billion.

  Circuit-Plus, a global ERP (enterprise resource planning) deployment for GE Lighting that reduced time and money spent.

  GE’s Business and General Aviation division’s advanced turboprop engine. Developed with FastWorks and an entrepreneurial team, the engine won a big deal from Textron against an incumbent. Time and again, I’d see Aviation as one of the most entrepreneurial divisions in GE.

  Preventing train derailments. Using FastWorks, the GE rail team reached across GE for access to technology that would help prevent storm-related derailments. An MVP was accomplished in ninety days.

  In the first year, eighty coaches were trained for almost one hundred global FastWorks projects, exposing almost one thousand GE executives to the methodology. As we expanded FastWorks over the next few years, that number would grow to nearly four hundred coaches and one thousand projects. Eventually, it simply became the way we worked. We even amplifed ways to use FastWorks every day. I loved hearing colleagues say things like, “What’s your hypothesis?” or “Let’s just MVP this.”

  The speed and depth of our success with our coaching approach was a major aha moment for me. Several simple tenets of innovation became clear: First, to build trust you need to stop viewing business units and project teams in terms of hierarchy and more in terms of local guilds or tribes—as communities of practice. Second, the emphasis isn’t on providing an answer so much as asking penetrating and probing questions that get people to reflect on the old assumptions and cultural habits that are influencing their thinking.

  I wish I had appreciated the role of coaches earlier—master craftsmen and -women who teach their apprentices the craft of imagination-making. Consultants tell you what to do and leave; coaches lean over a business leader’s shoulders and advise them: “Good. Good, bad. Right move. Now faster.” The coach’s role is the essence of emergence—it’s about setting a good “mission objective”—giving the teams the freedom to iterate and learn forward. Pretty soon, “mission-based teams” became pervasive ac
ross the organizations, with teams shedding hierarchy in pursuit of the outcome.

  As we led the rollout of FastWorks, it was clear that GE also needed to get rid of the institutional structures that inhibit these new behaviors. To create new beliefs and encourage new ways to act, we needed to rethink GE’s incentives and rewards.

  To this end, HR chief Susan Peters and HR FastWorks champion Janice Semper retooled the iconic performance management system—the way success and rewards were measured at GE since the 1970s. People involved in a process that demands productive failure cannot be judged solely by a system that rewards profit margin and punishes imperfection.

  Emergent leaders must be rewarded for the productivity of their failures—their idea generation, their pivots, their learnings, their clear-sighted zombie kills—as well as for their major successes. To do this, creating constant feedback loops—and getting rid of annual performance reviews—is absolutely critical. At GE, we moved to ongoing feedback via an app named PD@GE (Performance Development @ GE). Managers now give ongoing feedback to their employees—and vice versa—in a process that holds both accountable to each other and responsible for constant communication and coach-based behavior. Culture needs to be first in mind, not the last thing you do.

  By asking the right questions throughout the year, instead of annually feeding employees abstract benchmarks to hit, we were pivoting to an environment that gives permission, encourages candor, and sets up new levels of trust.

  I also developed something I called the Culture Club, a multilevel group in my business unit to give me feedback and, more specifically, be a way to drive culture change together. I needed our teams to hear directly from me that it was okay to test and learn, and they needed me to hear from them why it wasn’t so simple. We met quarterly. I challenged the team to bring me one thing I didn’t want to hear—something I or other leaders were doing that stood in the way of meaningful speed and change. I discovered a lot of time and energy was going into keeping reports, meetings, expectations alive that I had long forgotten about. We used these discussions as a way to reset expectations, drive candor, and hold one another accountable for change. Honestly, I wish I had done this much earlier.

  FailCon: How to Host a Convention of Failure

  The Culture Club identified fear of failure as a key issue to tackle. Our employees were good at their jobs, they cared, and they didn’t want to let their managers or their teammates down when something went wrong. They were failing at failing forward. So to get over that fear, they created FailCon, a failure convention to celebrate mistakes. Andy Goldberg, our creative director, led the way. We picked one day when everyone, in all of our teams, would share their failures in brown-bag discussions around the globe. I kicked the day off with a video sharing the Quirky failure and saying, “Let this be a testament to successful failure. We tried, we learned. New business got started in GE because of it.”

  It would take years for such ideas to be completely embedded in the company, but by the time the next stage of the rollout had been completed—in time for Jeff’s end-of-year deadline—it felt like we were rebooting GE’s OS, if not yet its soul. Those who had dismissed my ideas initially now not only accepted that there was value in this shift in our cultural OS—many even embraced it.

  Current

  In the spring of 2014, I headed west with Jeff for a “deep-dive review” of our digital operations at the GE Software Center in San Ramon. By this point, the home of GE Digital had a team of hundreds of software engineers, coders, and data scientists working on building the software that would surround our big iron machines with data and offer customers new outcomes to run their operations more efficiently.

  On the day we arrived, the team was delivering an update on “edge computing”—that is, computing power embedded in sensors and controls that allows computing operations inside industrial hardware as opposed to in the cloud. I got into a lively discussion with one of the center’s engineers, Tom DeMaria, who was explaining his excitement about the new wave of computing power that was coming to LED lighting. We were on the same wavelength. I knew it was only a matter of time that we (as in humans) would put amazing computing power anywhere there was a light fixture. Later that year, we planned to launch Link, GE’s first Internet-connected LED fixture for the home, developed with Quirky. I was thrilled by the idea of lighting as the connected infrastructure of a home, potentially replacing the need for broadband cable and creating a truly smart home.

  “Thank goodness we have a vibrant growing LED business,” I said. “The humble lightbulb drives the connected future, right?”

  “Sure, that’s great,” Tom said. “But think about the computing power once the Intels of the world start to focus their computing power on the Internet of Things. I know they’re already thinking about it, because we’ve been sharing our vision with them about the Industrial Internet.”

  He and I got carried away, riffing about the possibilities in connected buildings and cities.

  “With LED-based tech we could report on energy usage and help with energy management.”

  “Or provide location services. Maybe citywide Wi-Fi. Traffic monitoring and parking optimization.”

  “And then there’s security and safety. It could…warn about gas leaks.”

  “Or look out for terrorism. And that’s just the start.”

  Jeff eyed us with that look of “Focus here,” but he also looked intrigued.

  I rode with Jeff in the car to Menlo Park, where I asked him to meet with Sue and a few of our portfolio companies. (He always liked those meetings as a place to learn and also as a way to poke the business unit leaders about what they were missing.) “Wouldn’t it be funny if the oldest business at GE leads the way for our most advanced effort,” I said. “Think about it—we’re having a hard time getting people to understand what the Industrial Internet is about. Well, everyone understands lights. Every business and every home. That’s why I’m so keen on this connected-home project with Quirky.”

  “Sure, Beth. But you know we’re not a consumer company now,” he said.

  But Jeff saw enough opportunity here to make me an offer that caught me off guard: Why didn’t I oversee GE Lighting?

  “I like the spark of imagination you showed with the Quirky connected light project,” he said. “You have ideas for what to do with Lighting, and we need to do something different.”

  In a company as big as GE, Lighting wasn’t on many people’s minds. It meant a lot emotionally to the brand legacy of the company, but it was relatively small (for GE)—with $3 billion in revenue—and old and declining. But in it, I imagined the future.

  In the weeks after, I kept playing thought experiments on how we could put the LED unit on a path to greatness. Despite neglect—and downright hostility early on—the LED had become an almost $1 billion revenue business. Nonetheless, GE wasn’t interested in investing in a business that was largely built on a boring commodity, with declining profitability. While LEDs had a much higher price point, they also had a much longer shelf life—years, rather than months. Nonetheless, Lighting had shown the survival instincts of a determined runt. And it had grown the business as if its future lives depended on it.

  At the time, I was seeding a skunk works project with the Ventures team to imagine and map the future of energy. We were keen on energy storage as a way to make renewable energy scale—currently there is no affordable way to store excess sun or wind power at night or on cloudy days. But I was obsessed with solar, as we were with software that manages two-way electricity flow from consumers and businesses as they sought to gain more independence by generating their own electricity.

  We just knew that a more distributed energy world was going to drive the future, meaning that energy would increasingly be generated not only by centralized utilities, but onsite by businesses and even consumers. This made our utilities customers nervous, especially s
ince their path was fraught with regulation. But we could see the seeds of disruption emerging. A host of start-ups were coming into their own, and with GE Ventures, we were investing and partnering with many of them—such as STEM, a battery storage start-up run by a former GE marketer.

  As our Venture-led incubation took shape, I saw in Lighting’s LED-led resurrection a metaphor for reinvention. We would turn Lighting into the next generation energy company for the commercial and industrial space. By combining the energy efficiency of LED lighting with solar power and eventually storage capability, we could first save customers 20 to 40 percent on costs. Using software, connected through sensors to the LED fixture—we’d then deliver the future in an “intelligent” network in their stores, factories, and offices onto which other applications could be built and bring all kinds of new value.

  The story we created for the new business was simple and resonant: deliver unprecedented energy outcomes to our customers and provide them a ubiquitous digital network to run operations more productively. In a large retailer’s site, there are towering lampposts throughout the parking lot, and after the sun goes down they light the way for thousands of busy shoppers. But these are no ordinary lamps: they are high-powered LEDs, and embedded inside the lights are dozens of sensors and cameras. These intelligent light posts, well positioned high above the ground, have the ability to monitor and transmit information on available parking spots via a mobile app.