Overdosing on Lean Startup


Lean Startup by the book

Last november I embarked with another 3 people on a project called Prometheus. The whole team is pretty Lean savvy and so it came natural to start preparing a lean canvas, discussing value propositions, customer segments, building personas and coming up with hypothesis to test. This was pretty much Lean 101, by the book.

Because we're all remote (this will be the topic for another post), this process actually took several weeks to complete. I guess this should have already been a bit of a red flag, but we ran with it, after all exactly because we are remote, over communicating is a necessity and so it is being sure everybody has a chance to be heard.

Can it be too early to talk to customers?

Prometheus is focused on entrepreneurship education and after discussing potential segments we decided to start with transformational entrepreneurs, or in other words folks that are doing a startup to have a profound impact on the world.

We were just getting started defining this segment when someone very much like what we had in mind, let's call her S, showed up in our inbox.

It seemed like a great opportunity, but part of the group felt uneasy about it, which is understandable: we didn't have yet a clear set of problem hypothesis and working with S would have meant attacking the solution.

Rushing into solution validation and skipping the problem and channel validation is generally a bad idea: all of a sudden you find yourself stuck in a local maxima building something for one customer with no real learning and opportunity for growth.

We argued back and forth for a while and eventually decided to pass. That in hindsight was a big mistake.

Data, processes and comfort

Startups are by nature messy. There's a ton of uncertainty and most people aren't particularly good at dealing with it. In fact you may argue that we have evolved to err on the side of false negatives and avoid uncertainty.

So we look for external forces to ease the burden and the two most common ones are data and processes.

A lot of the young entrepreneurs that I've seen excited about Lean Startup ended up taking it too far, elevating it to a sort of magical recipe. But while it's tempting to write it off as inexperience, this is actually a very normal behavior that runs deep and is at the source of the thousands of courses and books that promise success in 3 easy steps: we all want the easy, quick, safe paths to our dreams. And gosh, why shouldn't we?

So when you have a group of people, especially if fairly new to each other, it makes a lot of sense to look for an external process to guide decisions and act as an objective arbiter.

What kills startups

I'm sure you've come across across this sentence before:

startups fail because they run out of money [or time, or both].

This is the argument at the basis of using something like Lean Startup: by testing your idea step by step you can limit the amount of waste and therefore increase your chances of success.

And it makes a lot of sense. If you spend all your money to build something nobody wants, you're done, no more playing around. If instead it takes you a few weeks and buying a few coffees/beers to talk to a bunch of potential customers to find out that nobody cares, you still have enough in your bank account to try something else.

It's about runway.

But as we've been harshly reminded that's only part of the story. The other huge factor responsible for startups failure is the founder[s] mental break down. This can happen for a bunch of reasons, but the biggest one seems to be lack of momentum.

Failure is not so bad if you can manage to get up and start climbing again shortly after you've fallen. But if you're stalling, even if you haven't fell, that just feels terrible. The number one problem I heard in my mentoring group is by far not growing as fast as expected (and this is another place the argument gut vs data creeps in).

Experimenting takes time and can be boring

Let's not hide behind a finger: running experiments can be incredibly tedious and emotionally wrecking.

Yes it's largely a useful framework to think through business problems, but that doesn't make it the most appealing thing to do every day. Lean Startup is often linked to the scientific method, but I'd challenge anybody saying that to spend a week in a lab tweaking experiments one variable at a time, going through the same routine day in day out even just for a week. It's not fun unless you have the kind of obsession some scientists do, and even then they also get discouraged, nobody is immune. Mind you, I'm not saying it's not worth it, just saying not pretty.

I've yet to meet somebody who got into startups for the fun of experimenting with businesses. Most people have a goal, something they care about, that they want to get done. Spending their days repeating experiments making little tweaks, waiting for lab results to come back, can be incredibly demotivating.

And that's what happened to us.

Because we wanted to do it right and be fair to everybody (remember, remoteness, doing it on the side, etc played a part in this), we decided not to work with S. A month later with just a bunch of canvases, personas and some hypothesis in our hands we all felt pretty bad.

Lessons learned and an example

The key thing that emerged from a discussion with the team was that we shouldn't just optimize for waste, i.e. test as much as possible to avoid doing stuff that won't produce results.

Rather, we agreed that we should optimize for momentum while keeping in mind the cost of execution.

In other words, if another S shows up in our inbox, we'll go ahead and work with her, keeping in mind that we need to funnel at least some of that work into validating hypothesis and whatever comes out of it should be just one data point, not the exclusive basis of our future decisions.

To apply this to a classic example let's take landing pages.

Up to a few months ago, if you came up to me telling me you had a brilliant idea and needed help building a landing page, I would have told you to forget about it and get on with talking to customers.

It's still true that there are many problems with starting with a landing page, primarily that if you don't understand your customers what are you even going to write as copy? And if you don't get clicks, what does it all mean? You don't know why it hasn't worked and therefore you've learned nothing, leaving you in the same exact position you were on day one minus the cash/time it took you to get that out.

Or not.

That's exactly the problem. On one hand, yes, there was no learning happening and that's bad, you're down in resources and nowhere closer to find a business model that works.


You probably felt good while building the landing page and excited about the prospect of having something out that you've made and that you can show to people.

Compare that to sitting through several conversations and putting together a customer archetype, structuring hypothesis and picking variables to measure. Then trying to structure an interview script, finding those people and finally getting in front of, or on the phone with, them.

The mental burden and reward of the two is almost at the two opposite sides of the spectrum. No wonder nobody wants to do customer interviews and everybody wants to build landing pages.

So what?

Conclusion: optimize for momentum and $$

Lean Startup or any other methodology does not exist in a vacuum, it's not right because it's right. Those approaches exist with the purpose of increasing your chances of success, whatever that means to you.

But the thing is, it's not just about learning as it's often portrayed. In Lean we stress the importance of learning as the goal of the build-measure-learn loop and any iterative based methodology will more or less focus on the same.

However when you look at the reality of spending a day as a founder, what's most likely to get you to the next day is not learning, it's momentum and excitement, passion. That's why investors look first and foremost for that in a founding team.

If we lose sight of our happiness and the value we hope to deliver we're guaranteed to not get to tomorrow. And then thing is, as much as learning is progress, for most people it won't be a source of joy in itself.

So here's my suggestion: if an idea come up that will take you weeks to implement and lots of resources think twice and probably figure out how to break it down and test it piece by piece; but if it's a matter of a few hours and it's gonna be fun, just do it and to hell with experimenting.

Enjoy your journey.

Join the community of Lean Startup practitioners

comments powered by Disqus