Is the Lean Startup Dying? On Failures, General Electric, and Bringing a Knife to a Gun Fight
If there was a Hype Cycle for all these methodologies and theories, you would find LS is sliding down the far end of the slope.
Steve Blank discussed the topic of Lean Startup being dead in his September 2018 blog post. If you’re crunched for time, here’s his conclusion:
In short, Lean was an answer to a specific startup problem at a specific time, one that most entrepreneurs still face and which ebbs and flows depending on capital markets. It’s a response to scarce capital, and when that constraint is loosened, it’s worth considering whether other approaches are superior.
Some people insisted this exception applied to the special 1% unicorns where capital was already abundant, or where corporations have the cash to burn. So for the other 99%, LS still makes sense.
For the most part, I agree, but I also don’t think it’s that binary. It’s easy to assume that LS is universally applicable to any problem and any industry because of the barrage of success stories by tech fundamentalists and the mainstream business press.
Like, what can’t LS help with, right?
Every methodology and pet theory gets their moment of fame. The excitement dies away once people see for themselves that LS isn’t always a fit for every situation. If you know a thing or two about Adaptive vs. Technical Problems; Simple Problems vs Complex Problems vs Wicked Problems, you’ll know what I mean.
Let’s talk about GE
Back in 2012, GE’s top management teamed up with Eric Ries who helped create the FastWorks program. GE was suffering as its old business model began to collapse. Much like the successful Six Sigma initiative brought on by previous CEO Jack Welch in 1995, CEO Jeffrey Immelt brought on the LS approach to GE’s product development.
The FastWorks program trained thousands of executives and launched more than a hundred projects around the world. GE transformed its product development, improved product quality, and hired outside industry experts to challenge the way it did things. Here’s an example of one of those projects:
An early example of FastWorks in action involved a proposal by GE’s engineers for a five-year, $500m project to upgrade its H-class gas turbine. Mr Immelt called for the new approach to be applied, starting with a proof-of-concept exercise costing $25m. The result, GE says, is an upgrade that should be ready in two years for half the original cost.
Eric Ries’s recent book “The Startup Way” celebrated the projects and successes achieved by the FastWorks program. Actually, Ries focused a lot on GE to drive his point home. It’s hard to argue against the results, like saving 80% in development costs.
Fast forward to the present. There have been massive layoffs and its stock plummeted. Jeffrey Immelt resigned in August 1, 2017. In June 2018, Bloomberg reported that GE was unseated from the Dow Jones Industrial Average after a stable 100 year legacy.
To be fair, GE struggled with an onslaught of past missteps and miscalculations in what other analysts call “questionable deal-making, needless complexity and murky accounting.” Not to mention, many of its other business units weren’t as transformative as GE originally presented them.
The tipping point began with the risky acquisition of Alstom, a large French competitor to GE’s largest business, GE Power, which forced GE to funnel insane amounts of money, time, and development on gas turbines. The FastWorks program was still in place. Fortune magazine reported that its hyperfocus on gas turbines would eventually hasten GE’s demise:
…Alstom’s profit margins were low, but GE figured it could raise them. GE’s strategy relied heavily on selling services, but regulators made the company divest Alstom’s service business. The acquisition added more than 30,000 high-cost employees, many in Europe, but GE figured they’d more than pay for themselves. Worst of all, the purchase was spectacularly mistimed. GE doubled down on fossil-fuel-fired turbines just as renewables were becoming cost competitive. Result: Global demand for GE Power’s products collapsed, while GE had bet heavily the other way. GE Power’s profit plunged 45%.
What was the real miscalculation?
All business models get disrupted eventually. Once you’ve sensed that much, how do you come up with the right strategy?
Some thought it was Immelt’s fiery belief in LS. After all, he did leave a complementary book blurb for Ries’s first book. Plus the lack of time to nurture and scale GE’s projects.
Steve Blank insisted FastWorks wasn’t the problem, but “activist investors” like Trian Partners who pressured ex-CEO Jeffrey Immelt into sacrificing long-term innovation efforts for short-term remedies to revive GE’s stock price. GE’s business units didn’t have the natural scale of software companies for rapid growth, so no innovation program was going to help its business units bounce back in a timely manner.
So, did someone bring a knife to a gun fight?
With Blank’s close ties with Ries, I’m not surprised his final analysis was understandably dialed down. But whether activist investors were present or not, if they had sensed this shift in the marketplace, would FastWorks still be the wisest option?
Because the bottom line was that the market was already shifting away from gas turbines to renewable energy. Eric Ries’s FastWorks program only helped GE improve and optimize on things that became less important over the years.
At this juncture, blaming LS as the single point of failure is irrelevant.
How long can you continue to just improve before a shift happens?
The good news is that LS generally works and is helpful. LS has incited entrepreneurial movements worldwide. It’d be unfair to dismiss this paradigm shift in business. After all, it has helped us solve problems better than ever before and pushed massive players your wouldn’t normally see finally adjust and catch up to the current marketplace. (Lookin’ at you government, academia, and defense)
The not-so-glowing news is that we’re now at an inflection point where we are now collectively witnessing limits to LS-style problem-solving. Those limits are usually deflected by those who — let’s say — have a deep stake in the LS “success narrative.“
This is not the first time the LS model for success failed to materialize, tough. And this is not just cherry-picking to prove a point. GE is just one publicized corporate case study out of smaller undocumented cases never to be written or mentioned for fear of embarrassment. More of these “failure outliers” will pile up until they can no longer be cast aside as mere outliers.
I’m reminded of a post back in 2006 (Yes, more than 10 years ago!) by a practitioner who had similar concerns when it came to Lean and Six Sigma which still rings true to me today:
Most of the business functions where Lean and Six Sigma concepts have worked have had a long-standing, well-defined business process already in place, and the implementation of Lean or Six Sigma or both have brought about tweaks or improvements to those existing processes. Is Lean or Six Sigma right for a business function or haphazard process that represents the state of the art for innovation in many businesses today? It seems to me that there are few documented processes that are followed repeatedly in innovation teams, and little consensus about how to innovate even within the same organization. Don't these issues need to be addressed first?
The punchline here is that there is a world of a difference — by orders of magnitude — between discovering the unknown versus optimizing what’s already known.
Learning from our mistakes should be encouraged. No argument there.
But not all failures are smart ones. Not all failures are based on an informed understanding of the battlefield. And not every misfire and failure will teach you what to do next. We shouldn’t default to failure as an honorable one every time a grand entrepreneurial attempt collapses. Many products and services that are predicated on the LS approach, by the way, end up being incremental improvements on what’s already been done. At worse, they’re ideas looking for a problem to solve.
We just need to be clear that whether DT, LS, or Six Sigma, any methodology you deploy will have a hit rate. Meaning: You’re gonna see some results of your efforts no matter what. But will those results get you closer to your north star? If you had a poor understanding of the problem and a crap strategy to boot, how far do you want to congratulate yourself?
Instead, clearly understand the innovation context: Are you trying to invent something new? Or do you want to consolidate and optimize the good things that you already have going on first? Would those decisions make sense in the next 3-5 years? Are you capable of carrying out this endeavor? Why or why not?
What is the root cause of the unmet need? Knock yourself out with customer feedback loops, the paradox with this question is that customers can’t always tell you what they want or need, nor will they have an explicit pulse on what’s really happening in the market.
So then what's next for Lean?
Mainstream business writers, copywriters, junior writers, and non-practitioners will keep doing what they do best: Stand guard for the next hottest thing.
Blank concludes that there's a next generation of Lean, and that is, the "Innovation Pipeline." He advocates a "self-regulating, evidence-based innovation pipeline" that prioritizes problems, ideas, and technologies. Lots of these keywords are interwoven with front end collaboration, and that’s a wide open field to explore another time.
At this point, how would you prioritize anything at all? All this situational awareness stuff needs to be pointed inward:
Do you understand your organization’s innovation priorities?
What opportunities do you see out there? How quickly can your organization act on them?
Have you identified where you need to innovate and what types of innovation you’re interested in?
Do you have a pulse on industrial trends and patterns?
If these questions are tough to answer, or you have vague answers, then it’s likely your subordinates will be in the same position, too.
The idea of a continual and disciplined problem-solving process that multidisciplinary teams can deploy and repeat isn't new. Actually, it spans 50+ years. Maybe further. So it’s unsurprising to see every flavor-of-the-month possess some similar carryover with its predecessors.
For example, we saw the same excitement when Design Thinking was just becoming a thing, which is really just creative problem solving at the core but with a branded emphasis on creating products, services, and apps.
If there's one thing you must take away from this, it's that you must prioritize sensemaking, situational awareness, and problem exploration. Again, it’s a wide open field. Only then will your chosen methodology, whether DT, LS, or Six Sigma, make sense with the overall strategy.
See you around when the next hottest thing gets hyped! ;)