A startup founder can suffer from two opposite diseases, and both are fatal. The first is stubbornness. The founder clings to the idea and drags it through the years, even though the market has long and politely been hinting that it does not need it. The second is thrashing. Every month the founder opens a new direction, because they read an article, went to a conference or talked to a single client, and in the end there are five unfinished products and not one that works. The advice "stay focused" cures the second disease but aggravates the first. The advice "stay flexible" cures the first but aggravates the second. I think the answer is not some golden mean but one simple image from basketball.
The pivot foot
The word pivot, so beloved by founders, came into business vocabulary from sport. In basketball a player holding the ball can turn in any direction, but one foot, the pivot foot, must stay where it is. If it leaves the floor, that is travelling, a violation. With the other foot you can step anywhere, look for a passing angle, get around a defender. This is the perfect model for a startup. Turning is possible, and it is normal, but only while the pivot foot stays planted. Lift it and it is no longer a pivot but an escape, and you have to start from zero.
What is the pivot foot in business? It is what does not change when you turn: a deep understanding of a particular problem, a particular audience, a particular technology. Everything else can change. A classic example: Slack started as a company making an online game. The game did not take off, but the internal communication tool the team had built for itself turned out to be worth more than the game. The pivot foot was the ability to make convenient tools for people communicating online, and it stayed put while the product turned.
Scattering looks like flexibility
The most dangerous thing about scattering is that from the inside it feels like energetic entrepreneurial flexibility. We are building a service for beauty salons. Then an acquaintance from the fitness world shows up, so let's do one for fitness too. Then an investor asks about a mobile app, so let's do mobile. Then someone says B2B is more profitable, so let's do an enterprise version. Each decision on its own is logical. Together they turn a team of five people into five teams of one, and none of them finishes anything.
A founder once came to me with a plan that had several products for different markets scheduled for the first year. I asked which of them he would close last if he had to close all but one. He thought for a long time and then named the one that was already bringing in money. We agreed that the others would become not products but hypotheses in a notebook, to be revisited once the first one worked without his daily involvement. A few months later he admitted that for the first time in ages he was getting enough sleep.
Focus is not the ability to say yes to what matters. It is the ability to say no to what matters less, even when it is very beautiful.
How to tell difficulties from a dead end
The hardest question is when to turn and when to hold on. There are always difficulties, and if you turn at every one of them you will never build anything. But dead ends are real too. I use a few rules.
- Criteria are set in advance. Before an experiment we write down: if we do not get such-and-such a result within such-and-such a time, we change direction. A decision made in advance is more honest than one made in a moment of fatigue or euphoria.
- Every experiment has a deadline. An open-ended experiment is just a new direction quietly eating resources.
- Watch behaviour, not words. If customers say "interesting" but do not pay and do not come back, that is a signal. If they pay but complain, that is a good sign: the product is needed, it just needs improving.
- Execution problems are cured by work, model problems are not. If you lack hands, speed or quality, hold on and keep working. If people do not want what you offer, no amount of work will change the answer.
Architecture that allows a pivot
There is a technical side to this topic that rarely gets discussed. Many startups cannot pivot not because they do not want to but because their system will not let them. In a year they have written so much code for the original idea that any change of direction means rewriting half of it. The code becomes cement in which the pivot foot is stuck, except now both feet are stuck.
That is why at an early stage I always advise building systems that are cheap to change. At Ingello we use Fractal for this: the system is described and the generator produces the code, the database and the interfaces. If the direction changes, we change the description and generate the system again instead of taking apart a monolith that took years to glue together. Incidentally, Ingello itself has a pivot foot too. Systems, our engineering bureau, earns the money, and the other directions live on that money. We can experiment with anything as long as the pivot foot stays planted and feeds the family.
What to do tomorrow
Take a sheet of paper and draw two columns. On the left, what does not change for you: the problem you solve, the people you work for, the knowledge you have that others do not. On the right, everything else: the product, the sales channel, the pricing model, the technology. The left column is your pivot foot, do not touch it. The right column is your other foot, you can and should step with it. Then look at how many directions you currently have open at once, and close everything that is not connected to the left column.
