🧨 The Mean Startup, how GE failed at doing innovation by the book

🧨 The Mean Startup, how GE failed at doing innovation by the book
Photo by Thomas Kelley / Unsplash

General Electric (GE) was founded right after the American civil war at the early stage of the industrial revolution. Today, with around 300,000 employees working worldwide in a dozen markets, from home appliances to energy and aircraft engines, GE is still a force to be reckoned with. Think of it as a Google that would have emerged from the birth of electricity, not the internet. But since 2008, GE's market leadership steadily declined, making it the worst stock on the Dow Jones index constantly. This decline devolved abruptly in 2016 with the retirement of its CEO, Jeffrey IMMELT, a market cap divided by two while the stock market was up 41%, and a cascade of divestments (some of them to the Chinese HAIER — a sign of times). Since then, while GE jet engines, medical scanners, and power plants are still states of the art, the company went into a spiraling free fall.

What will be interesting to us in the middle of all this is that GE did all the right things in terms of innovation. Or at least all the things that innovation gurus recommended as the best practices.

Startup Envy

Jeffrey IMMELT had an epiphany six years ago in the middle of this turmoil. A clear strategic vision to reboot this falling giant. One that is shared by many CEOs right now but without the opportunity to do it at scale:

«Let's put back the customer at the center of the business. »


« By acting more like a startup. »

The way to reboot profits globally would come from « lean startup » principles. Since 2015 this was the Harvard Business Review-grade solution to understand digital, regain competitiveness, and survive the next century.

« Starting in 2009, over the course of several years I visited our controls and analytics labs and spent time in Silicon Valley. As the CEO of GE, I could get the best people in a field to talk to me about what was going on. I always capitalized on that. I met with tech leaders including Jeff Bezos, of Amazon; Paul Otellini, of Intel; Marc Benioff, of Salesforce; and Steve Ballmer (and later, Satya Nadella), of Microsoft, and had dinners with venture capitalists. I listened to them describe where they were going and how they went from strength to strength. I also read a lot. The two things that influenced me the most were Marc Andreessen’s 2011 Wall Street Journal article, “Why Software Is Eating the World,” and The Lean Startup—Eric Ries’s book, which I literally read in a day. » Jeffrey R. IMMELT in HBR, Sept. 2017

Does it ring a bell in your own company?

Or, and don't get confused if in your company you're saying « design thinking » or « open innovation,» these are all variations of this epiphany.

The real difference is that your board makes a show of it and won't allow you to do it at scale. Past a few workshops where post-its are aggressively used by executives and new business models are invented on the fly, no one intends to do lean, design thinking, or whatever. Worst case scenario, you build a flagship innovation center and have a few weeks of press releases. Then it's back to business as usual.

Not GE.

With nothing much to lose and still a lot of cash and assets, General Electric decided to go full steam and be serious about it. The press, financial analysts, and keynote speakers loved it. Was the motto « Lean startup? If GE can do it, you sure can! »

And as history just explained to us, this was an industrial failure.

Let's try to understand why…

To Be Fair

First of all, let's not shift the entire blame for GE's fall from grace on the lean startup methodology.

Many root causes are apparent: an ongoing lack of vision and strategy, an old incumbent culture, unpredictable market shifts, geopolitical turbulences… But more precisely, under Jeffrey IMMELT's leadership, GE went through a spectacular string of terrible acquisitions: Telemundo, Amersham, WMC Mortgage, Alstom… On top of that, the company managed to get hit by the US Securities and Exchange Commission for « creative » accounting and suffered from it quite a lot (as they should have).

This is quintessential to understanding the demise of GE as an economic leader. But if the problems were numerous, their ambition to develop a transversal startup mindset didn't help and probably just accelerated this whole mess.

The Scale Conundrum

One of the core beliefs of GE's CEO was that « You can't regard a transformation as an experiment. »

We’ve approached digital very differently from the way other industrial and consumer products companies have. Most say, “We’ll take an equity stake in a digital start-up, and that is our strategy.” To my mind, that’s dabbling. I wanted to get enough scale fast enough to make it meaningful. My view was that GE had as good a chance as anybody at winning in the industrial internet, because we were not starting from scratch: We had a $240 billion installed base of service contracts, a huge order backlog, and the ability to offer financing. We could build on our existing strengths to get even better. So we launched digital across all our businesses. By that I mean we launched a major effort to embed sensors in our products and build an analytics capability to help our customers learn from the data that the sensors generated.Jeffrey R. IMMELT in HBR, Sept. 2017

See, this is very lucid.

You're not going to reshape GE with, say, a corporate incubator churning through a dozen intrapreneurship projects a year or by partnering with a VC fund to spray a few million on a few more startups. Can you give me an example of any corporation that found its new business with such an approach these last five or ten years? (No, you can't.)

That being said, building up peripheral capabilities to be more customer-driven and lean could make sense to test the waters. And then you do need to build at scale. That much I agree with. For instance, a solid VC portfolio would need more than a hundred investments over two to three years to record success. Innovation is a volume game. This is why you need scale.

In the end, the requirement for scale is driven by 1) the need to explore a large volume of uncertain options before unmasking real future opportunities and 2) reaching a critical mass of feedback for each exploration to decide something.

Let's consider a company like GE through a heat map of uncertainties :

You see here how you would organize teams with different KPIs and mindsets about innovation.

The « red shirts » in Star Trek are sent to the most dangerous zones away from the Enterprise. For business, it's where the company's future is building up and yet wholly uncertain. The red shirts that explore new technologies, you call them R&D. The ones that explore recent market trends… you usually don't have them (and no, it's not the marketing which is generally a yellow team at best).

red shirts star trek

But the red shirts can't be made accountable to financial KPIs. There is no business plan where they are. If they can write one, they play too safe and are not doing their job. Their job is to map the future before it's explained to you by Gartner or PWC.

If they don't aim at bringing money back directly to the company, they still bring different and essential ROIs to the core business:

  • Sorting attractive future market opportunities from fads and hype (market intelligence)  ;
  • Getting to understand how to unlock the potential of new technology and if there is Moore's law of sorts at play (technology scouting) ;
  • Identifying new capabilities to simplify and enhance current processes (process agility) ;
  • Growing unique cross-team expertise and accelerating best practices spread (cultural agility).

If you're not dissociating these different objectives and allowing them to flow back to your core business in a controlled way, you're asking your core business to take crazy pills. Which for GE translated these last years in :

  • Let's put giant touch screens on ovens because of Millenials!
  • Let's have biometrics sensors on faucets because they're becoming cheap enough!
  • Let's 3D print food because it's the future!

Exploring these ideas makes sense (a lot of sense). But thinking they'll get to the market demonstrates a deeply flawed understanding of the market.

For instance, if you're GE, the Millenials are not your market yet. They don't have purchasing power; they don't have families… Hell, most of them are still living with their parents. When they're ready to buy home appliances, the market will have changed. They will have changed. Older Millenials tomorrow won't think and purchase as young Millenials today. So getting an idea of how they feel today is interesting for sure but doesn't linearly translate into products to be presented at the CES Las Vegas.

There is a very millimetric quality to how you manage the cold, no-nonsense zone of a business and the red hot let's-go-crazy fringes. Removing friction and letting all that be the same process with the same people and mindset is kind of recruiting train drivers to race on Formula 1 and vice-versa. Who would do that?

But also, if you want to be ultimately customer-driven, then you have to focus all capabilities on delivering what the market asks right now. Not in ten years, not in five or two. Right now. This is not a logic compatible with innovation. Are you in the mobile market? Then deliver an iPhone look-alike with a notch that costs less than Apple's flagship product. Sounds dumb? It sure is. It's what the market wants, though. So what is it going to be: being innovative or market-driven?

Do you want both now? OK, sure. But not at the same time with the same teams, or your company will be torn apart. And then, rapidly decide who is leading in the end: innovation (and like Apple, you decide that the customer doesn't know what he wants) or customers (and as L'Oréal, you want to deliver as fast as possible)?

If you say, as Jeffrey IMMELT, all that is put back in perspective, that the first pitfall at being « more innovative » is managing the innovation scale, I would wholeheartedly agree.

I don't know why GE would want to infuse « startup logic » across all teams and departments. Hot uncertainty zones and cold ones don't work in the same way.

So maybe GE started to drink too much of the startup Kool-Aid and started to believe they were a 12-person team and not 125 years incumbent with an extensive legacy business. Maybe in 2013, Eric RIES sold more than he conceptually had in stock. Or perhaps the idea was just too lovely to resist. We will neither know, and to be honest; it's not that interesting.

Interestingly, the cautionary tale that GE represents is when innovation tools seem so lovely that you apply them irresponsibly all across the board. The correct scale at « being a startup » is about « red shirt » teams being able to test enough opportunities with enough market potential to probe the future and then have other groups that, with this knowledge, decide where to nudge your core business before it's too late progressively.

There's our first obvious post-mortem right there:

If scale matters in innovation, it doesn't mean that everyone should get involved.