Deliberate decisions.

Learning needs something to compare against

A decision is a claim about the future. We’re doing it this way, because we think it will give us that.

When the outcome then arrives, you can hold it up against the claim, and only then does learning happen. Did it turn out the way you thought? Did it turn out better, and if so why? Did it turn out worse, and did you miss something you actually could have seen?

If the thing merely happened, on the other hand, there’s no claim to hold it up against. You have an outcome and nothing to measure it against, and then the conclusion is either none at all or a guess. That’s why companies can do the same thing for ten years without getting better at it. Experience doesn’t arise from things happening, it arises from someone comparing what happened with what was intended.

It holds just as much when things go well

This is where most people lose the thread, and that’s why this deserves a text of its own.

Nearly all the attention goes to decisions that went badly. What went well is rarely examined, because “well, it went fine” ends the conversation. But a successful outcome nobody understands is exactly as useless as a failed one, for the same reason: it can’t be repeated.

You won a big deal. Was it the price, the person, the timing or the reference? Without a written-down assumption beforehand, the answer is constructed after the fact, and next time you carry the wrong lesson into the quote.

A company that only performs post-mortems on its losses learns to avoid things. A company that also examines its wins learns to do things.

Luck and skill look the same in hindsight

There’s no way around this. In hindsight, a successful decision and a lucky break are impossible to tell apart, if nobody wrote down what they thought would happen.

That isn’t an academic point. It decides what you put more money into next year. A company that confuses luck with skill doubles down on the wrong thing, and only notices when the money has run out.

Written-down assumptions are the only known cure, and they don’t have to be sophisticated. “We think this will get us three new customers by autumn” is enough. When autumn comes you know something you otherwise wouldn’t have known.

Decisions you make because it feels good right now

Then there’s the uncomfortable sort, of course, the one that solves the problem now and creates a bigger one later.

A customer is taken on who doesn’t belong in the strategy, because the calendar is empty. An exception is made on price so the deal closes this month. A system is bought because it exists today, not because it will hold up in two years. A person is hired to fill a hole, not to solve the task.

Sometimes such decisions are entirely right. It’s often better to do something than nothing, since a half-good decision made in time produces movement, and movement can be corrected. A decision never made gives you nothing to correct and costs you anyway, just more quietly.

The problem is that they’re rarely made as decisions. They’re made as relief, and that’s something else. Nobody asks what it costs in a year, who’ll have to handle it, and what we do then. There isn’t time for the question, and it feels uncomfortable besides, when everyone wants to move on.

A deliberately bad decision and a bad decision look the same in hindsight. The difference doesn’t lie in the outcome, it lies in whether someone factored it in.

The person who has to clean up

There’s one more reason, and it isn’t about governance but about people.

I’ve sat many times cleaning up after decisions that were made long before I got there. All you see is the mess, and it’s hard not to think badly of whoever caused it. Who on earth thought this.

If I’d found three lines about why the decision was made, the task would have felt completely different. Not easier, but more reasonable. It usually turns out that the person who decided knew exactly what it would cost and did it anyway, for reasons that were valid at the time.

The note gives the person deciding a thought to think through to the end. Whoever inherits it gets something more: an explanation, and with it the ability to judge whether the reason still holds. In a company that’s going to grow, the latter is the whole point, since almost everything anyone does five years from now rests on decisions they weren’t part of themselves.

What the note needs to contain

Not much. Five lines, written the same day the decision is made.

What we chose. One sentence, in plain language.

What we chose against. This line is what makes the note a decision rather than a jotting.

Why. What would work thanks to it, concretely: this deal, this delivery, this month’s liquidity.

What we think will happen. Both the good you’re counting on and the price you think it carries. This is the line that makes the note a lesson rather than a defence speech, and it should be written even when the decision feels obvious.

When we’ll take it up again. A date. Without a date it isn’t a postponed decision, it’s a forgotten one, and no comparison ever gets made.

The last two lines are the whole mechanism. One is the claim, the other is the day you correct it.

Something to try

Correct an old assumption. Take a decision you made about a year ago and write down what you think you believed then. Compare it with what happened. Uncomfortable, and almost always instructive.

Post-mortem a win. Next time something goes really well, take twenty minutes at the next meeting and ask what actually caused it, and how you know. If nobody can answer, you’ve found a gap that costs you more than any loss does.

Ask what just came to be. Go through customers, services, prices and routines and put one question to each: did we choose this, or did it just turn out that way? The customer base is the most common place, since nearly every company has customers taken on when the calendar was empty who have then come along for years. Pricing is the second most common: a price set seven years ago and then adjusted by an index isn’t a decision, it’s a habit with decimals.

Go looking for an inheritance. Something you’re annoyed by today that somebody else decided. Find out why. Either you’ll find a reason that still applies, or you’ve just been given permission to change it.

In practice

In Genomlysningen this is often the most uncomfortable part and the most liberating. Uncomfortable because the list of things that just came to be is almost always longer than the owner thought. Liberating because everything on that list can be changed without anyone having to have been wrong.

What makes the most difference over time costs nothing. Add a final line to every set of minutes where something is settled: we’re doing it this way because, we think it will give us, and we’ll look at it again then. After a year you’ll have two things you didn’t have before. A history that explains the company to everyone who comes after, and a collection of assumptions to correct. The second is what makes you decide better next year than this year, and there’s no shortcut to it.

A deliberate no is a strategy. An unthinking yes is just what happened to happen.