Is the startup industry suffering from bankruptcy bias?

This post from Eva de Mol made me think about the graph below which I presented a lot during my first months at YES!Delft when I trained Validation Lab teams. The chart comes from CB Insights and you can find it here. It shows the top 20 reasons startups fail. It is hard to point your finger at one reason your startup failed, and that’s why CB Insights did a post-mortem analysis on 101 startups. They asked the founders why they thought their startup went bankrupt and everyone could check the boxes (so they could check more than one box, that’s why it doesn’t add up to 100%).

Bankruptcy reasons

At YES!Delft a lot of our support was consciously or unconsciously geared towards solving these bankruptcy reasons. Validation Lab to fix the market need, early pilots through Innocubes and financial advise through YES!Funded. YES!Talent to fix your team issues, a lot of design thinking in our early programs to make sure the product would be user friendly, and this list goes on and on. What I’ve seen from other startup support organizations in the Netherlands is that we were not alone in this way of thinking.

When you look at the content and coaching of the programs you could say that incubators and accelerator programs are geared towards preventing bankruptcy. And YES!Delft was fairly successful in this, we used to say that almost 80% of the deeptech startups in the ecosystem were successful (where successful was defined by: survived or made an exit). In other words: not going bankrupt was one of our definitions of success.

Not going bankrupt is not the same as having success

Here is the thing: not going bankrupt is not the same as being successful. Actually, there is research done by Paul D. Reynolds which shows that from a society point of view the worst you can get is an “active startup” which is so much as a startup which survives, but is not making a profit (anytime soon), is not growing (fast), is not developing new products but is also not on the brink of bankruptcy. Why are these so bad? They are not creating (a lot of) jobs and they are not providing innovation, yet they are using office space, applying for grants and subsidies and worst of all: the opportunity cost of their founders and employees are extremely high: imagine the amount of energy that goes into running a company, what if that energy would go somewhere else, somewhere successful, imagine the effect of all their energy in a company which does move towards success!

In my own words Reynolds pushes for having profitable businesses as soon as possible (this is not the same as actually making a profit, startups hardly make a profit in the first decade) or quit the startup as soon as possible and look for a new idea, a new opportunity or a new job. So, what does that mean for the startup industry? Instead of supporting startups to not go bankrupt in those early stages, we should encourage them to quit as soon as possible?

No, not really. Success is hard to predict in early stages and sometimes people want to quit while they shouldn’t: the team is not yet complete, the idea needs pivoting, there is no business model yet, etc. etc. Although it takes a relatively long time for startups to quit it is not always the right thing to do.

Survivorship bias

I can see why your confused right now. We should not create some sort of soft cuddly environment where literally no one goes bankrupt, but as we can’t predict success in early stages what should we do instead?

The first thing we should do is to get rid of our bankruptcy bias. We should do the opposite of what people who suffered from survivorship bias did.

Back in WOII these were the planes which made it back. The red dots are bullet holes, so it made sense to enforce the points which are almost all red, right? Well yes, if you focus on the planes which made it back (the survivors).

“The bullet holes in the returning aircraft, then, represented areas where a bomber could take damage and still fly well enough to return safely to base. Thus, Wald proposed that the Navy reinforce areas where the returning aircraft were unscathed,since those were the areas that, if hit, would cause the plane to be lost.”

If you think about it, the engineers had relatively easy time. There were just two categories: the ones who survived and came back, and the ones who didn’t. You can’t check the ones who didn’t come back, so you take the reverse of the ones who survived, and you strengthen the white parts of the planes, not the red parts.

Bankruptcy bias

Unfortunately, success is not as clearly defined as a surviving airplane. We have a hard time looking for successful startups, not because there aren’t any, but because success is something different for everyone involved. It’s not black and white, it’s black and all shades of grey.

Because we haven’t yet defined startup success, let alone defined the factors influencing it, it is much easier to analyze the startups which are bankrupt. At least there is no discussion on the definition, and as long as we do the opposite of what the bankrupted startups do, we’re all good, right?

So, we’re analysing startups which went bankrupt, we’re pulling them apart, having long interviews with founders and draw our conclusions. A lack of customers? Let’s setup a customer validation program to give them customers. A lack of funding? Lets spend money! A lack of focus? Let’s give them a mindfulness course and free yoga. We’re actively strengthening the white parts of the planes and the result is that we have a lot more startups in the air than a few decades ago, but are those startups really creating wealth? Are they having an impact on the world or are they mostly just flying around? Shouldn’t these pilots be on the ground? Doing something else with their lives?

How to define startup success

When we shrug of our bankruptcy bias we should think about why we are so focused on “the losers” of the industry. It doesn’t make sense to spend so much time and energy on the ones that come in last, the ones that get knocked out in round 1, or the ones that don’t even make it to the finals. I think part of the problem is that we lack a clear definition of a winner, so if we want to take a look at the winners and find out how they did it, we don’t even know where to look! Sure, you can name a few startup successes (usually the unicorns we all know), but are they really a success? Facebook? Uber? Amazon? Are those the companies that we want more of? Do we really need more unicorns? Should we copy those blueprints and act on them? I’m not so sure.

Where we should start

Yes, there are a lot of lessons from studing bankrupt startups, and we should definitely act on them. But we should not expect that this is enough. Not going bankrupt is not the same as having success.

As the definition of success is already unclear we should spend much more of our energy on redefining what success actually means, and go from there. It’s OK if this is not the same for everyone. This definition will change for municipalities (jobs, jobs, jobs), for commercial VCs, for public incubators, for universities, and so on. But we should talk about it, and not just assume we all want the same. What are we aiming for as a startup support industry? Is it really just 1 billion dollars of revenue, or is it more?

Team team team

No matter the definition of your success, because of my experience at YES!Delft, I can already predict one of the successfactors: the team is much, much more important for success than we currently give them credit for. This is not real news though, VCs have been telling us for ages that the team is the key to success. Yet, most of us are still selecting startups based on their TAM, their IP, their traction, the product, or their signed pilot agreements instead of their ability to form the best team imaginable. This also shows in the way we talk about teams. While we have numerous definitions and criteria for market, technology, and funding, we’ve just started defining what a good team actually is (a hacker, a hustler, and a hipster, come on..), let alone that we’re able to predict their team dynamics.

Let’s start with defining what we actually want, what successful startups actually are and go from there. In the mean time we should already try to talk more about teams, what makes them good and what makes people stay at those teams.

Startups, scale-ups, accelerators, incubators, business builders, venture growers & ecosystems. Dutchie living in Sweden.