A few months back, I gave a tech-talk at Pivotal Labs on the reasons why people fail to implement lean startup ideas, even when they like the theory. I’ve been meaning to post the notes to the slides, so here you go:
When “lean startup” first came out, it was greeted by two erroneous responses:
- “this is a roadmap to guaranteed success!”
- “thanks Captain Obvious, but we already do this”
However, lean is not:
- asking customers what they want (or if they think your idea is cool)
- simply slapping the term “MVP” on your release
- trying a new idea every 10 minutes
- doing things cheaply
So how do I think about lean, especially at the early stage?
- you frame your business as a set of hypotheses
- you get out of the building (and your own head)
- you ruthlessly challenging your ideas via focused experiments
- you figure out the lightest, fastest way to build confidence in your idea
- you prioritize validated learning over product and growth
- you run fast iterations that end with validation, not acceptance
- you measure progress and factor that data into your decisions
All of which are tactics to de-risk your innovation ambitions with maximum efficiency.
So why is it so hard to do?
It is really hard to balance the reality distortion field with ruthless questioning. You need tons of confidence to attract teammates and capital, and yet somehow you need to be massively humble and challenge yourself vigorously.
Vision is fun, and it is what drives us as founders and innovators. No one likes learning that they are wrong. It is human nature to procrastinate bad news. When you are happily designing and building product in the ivory tower, your vision is going to be 100% right, your customers are 100% happy (because there are none), and your business potential is untarnished. Heck, you can sometimes get a higher valuation from investors if you haven’t actually hit the market (although usually only previous-exit entrepreneurs can pull that off).
And of course, there is also the fantasy of being Steve Jobs, listening only to your gut (nevermind the incredible amount of resources and talent that Jobs gets to throw at a problem).
This is a tricky one. Domain expertise can be a critical edge for a startup. Unfortunately, too many domain experts view their knowledge and experience as proven facts, not hypotheses. If and when you are innovating, you have to be willing to question what you know because you are creating a new context by the very nature of innovating.
However, it is hard to admit that your expertise can also be a liability. The critical thing to remember is that innovation means that you are guessing.
Our industry lionizes getting funded. A funding event is considered success, and can give an entrepreneur rock star status for at least a little while. That can be a good or bad thing for a startup depending on how well they capitalize on it vs get distracted by it. The key thing to remember is that funding gives a startup needed oxygen, but it is not the same thing as true success.
VCs want to see growth, but don’t always question how sustainable or meaningful it is. Lean discipline can fall by the wayside as companies try to manufacture growth, rather than use lean to focus on fundamental progress, non-vanity metrics, and sustainable success.
Too often the entrepreneur is demanding “build my shiny thing!” However, lean values learning over product.
It is natural to hunger for your vision, your solution, to be realized right away. In many ways, lean says “be patient, be incremental, be real”. Patience is hard. And to the software developer, shipping product can feel like progress but we all have to remember that it is not.
(This point #5 is the most dangerous to me because I am a “maker” by nature. I love to design and build things. It’s fun to build first, and then later see what happens. Of course, the bigger the vision, the more you can convince yourself to build. Lean startup brings me useful discipline.)
The startup world thrives on “survivor bias” stories. But startup-land is like an iceberg, and as dangerous. We see a few successes sticking up above the water and try to copy their actions and their perseverance. We ignore the vast volume of startups under the water who failed regardless of their perseverance and vision.
People often succumb to “all or nothing” product design. In other words, why bother testing something partially-done because it will inevitably fail? The answer, of course, is so that you can learn. The second thing to remember is that if you need all the pieces of a puzzle to fall into place perfectly in order to spot and then achieve success, then the odds are seriously stacked against you (even more than normal).
Lean startup takes scientific-method methods, but it is not pure. Variables will be tough to isolate. Results data will be unclear. Sample sizes will be small. It is easy to say, “since I’m having to use my judgement and gut anyway, why bother with all this testing?”
With lean, you are putting yourself, your ideas, your reputation out on a limb. You are shoving yourself in front of customers knowing you might fail. Ultimately if you are being rigorous, there’s nowhere to hide.
The first rule of lean startup is that there are no rules to lean startup. Every innovation has its own context. Priorities and tactics will differ hugely. That makes lean startup difficult to grab onto and run with.
Ultimately, there is a lot of human psychology that fights the effective and continual adoption of lean startup, but the counterbalancing forces are fear of wasted time and a deep respect for how hard startups really are. But some people might need to get burned once or twice first before they really realize it.