The Gap a Mistake Opens

Updated 3rd Aug 2026

Learning from mistakes: beyond 'fail faster'

I love a good conference, and at this one in Romania, I'd found a delightful seat and was enjoying a pleasant coffee ready for the keynote of the day. The speaker put a phrase up on the screen in letters about a 25 metre high, in a nice font – I may be exaggerating – about the size, not the font.

FAIL FASTER.

The room nodded. Of course it did. It's a tried and tested phrase. It's a good phrase. Energetic. Slightly reckless. A phrase that looks particularly good behind someone wearing trainers with a suit on a conference stage.

I've always thought "fail faster" was about half an idea. Failing faster is only useful if you pay attention to what happened. Otherwise you're just breaking things at speed, and in most cultures that's frowned upon. And not all mistakes are equal, either. Some are lessons. Some are scars. Some end careers and businesses. So speed is not the point.

Every mistake opens a gap: the distance between what you expected to happen and what actually did. Inside it sits something remarkably valuable if you choose to look — evidence, insight, a clear view of reality rather than how you'd assumed things should be. But only if someone takes the time to see, digest and understand what's in that gap. Sweep the mistake aside and close the gap quickly, and the value goes with it. That's the difference, over ten years, between a team that compounds into a brilliant team and a team that keeps repeating the same mistake each year.

Now, isn't that worth pausing on for a minute. Because most of us treat a mistake as a thing to get past as quickly and quietly as possible — and in doing so we throw away the most important piece of information we're ever handed from reality.

Early in my career I fell in with a startup that broke things constantly. Not through carelessness — through relentless careful curiosity – a habit that I like to think I took into the company and one that's stayed with me since. Every time something went wrong, instead of asking who, we asked why. Then why again. And again — five times usually, sometimes more, until the answer stopped being a person and became a system problem we could point at.

We ran hundreds of these. The point was never really the cause. It was the valuable question hiding behind it: how could we have prevented this, caught it sooner, or built the thing so it couldn't happen in the first place? A broken afternoon became a permanent improvement, usually several. That trade — a bad hour for a better system — is a wonderful deal on offer in working life.

Which is why the absence of mistakes is so rarely the good sign we take it for. More often it's decision paralysis, or, we're not really tacking anything truly meaningful to begin with — a person, or a team, too worried about being wrong to put anything real at stake. Growth is a creative act; creative acts carry risk; risk produces mistakes. Mistakes are the likely outcome on doing anything that matters. The question was never whether they'll happen – they will – it's what you do with them when they do.

And here's the part worth mulling over with a nice coffee, because it runs against every instinct: most mistakes aren't really about the person who made them. People don't start the day wanting to get things wrong — most of us dislike it intensely, and we always feel bad making mistakes. So when it happens, look at what was around them. Attention scattered due to too much work being asked of people. Meaning clear in one head and no-one else's. Information missing, incorrect tools being provided or pressure to deliver at all costs. Expectations that never quite matched reality. No feedback until it was too late. These aren't character flaws. They're design. The reflex to find the fault in a person is nearly always the lazy person's version of the harder, more useful work of finding it in the system — which is, nicely, exactly where the fix itself lives too.

So when a mistake surfaces — yours, or someone else's — the useful moves are dull and steadying and rather unglamorous. Gather the facts. Understand the context. Contain the damage. Say clearly what happened. And then, the part that actually compounds over time: learn from it, and change the system so that particular mistake can't happen in the same way again.

I've found that this is the better route, rather than what I've seen play out in countless places; humiliation of those involved, panic, or blame someone else to avoid looking bad.

A decent rule of thumb that holds up through life: respond to another person's mistake the way you'd want yours responded to. Curiosity, not control. What happened, and why, and what we can improve — not who should be embarrassed or who needs an emergency performance review.

Watch what a mistake does to a place, though, and you'll usually see the opposite. Control tightens. Reporting becomes watermelon in shape and texture. New approvals appear. Another gate, another sign-off, another layer of oversight.

Sometimes that's genuinely needed. Often a form of self sabotage masked under process improvement. A process added not because it'll prevent the next mistake, but because it makes someone feel that action has been taken. Process can prevent a catastrophe. It can also smother the very learning you needed, not to mention our ability to move ideas into something valuable.

A small rule of thumb helps here, and I reach for it often. The first time something goes wrong, it's a lesson. The second time, it's a warning. The third time, it's a pattern. Lessons and warnings belong to the moment they happen in — you learn, you adjust, you carry on. Patterns are different: a pattern is the system telling you it needs changing and improving. This might look like better training, better tools, clearer ownership, a redesigned workflow and more systematic view of how ideas becomes value. Nearly all overreaction comes from muddling these two. Treat a single mistake as a pattern and you get the cascade of process that kills experimentation, creativity, flow and innovation. Treat a real pattern as a single mistake and you get the same mistake, probably in various different forms, forever.

None of this works in a climate of fear, because fear hides information, and hidden information only makes problems harder to solve. A person who's afraid to raise a mistake early hands you the larger bill later, with interest and a summons.

Which is also why "we want innovation" and "we punish the things that precede it" can't both be true at the same time. Innovation without the real possibility of getting it wrong isn't innovation. It's a slot on the calendar or nice words on a wall, or a virtue signalling attempt to make people feel valued.

Of course, this could all be seen as recklessness, to make mistakes all the time in the name of creativity and innovation - and that is not my intention. So, here are two rules of thumb I've held on to for a long time.

Responsibility should match experience: a new starter shouldn't be alone on the live system or deciding on a million pound contract; someone new to a process shouldn't be making the irreversible call without a hand nearby to guide. The second move is to build sandboxes, experiments and safe contexts. Places and moments where people can stretch their decision making and get it wrong without breaking themselves or the business. That's not coddling people or taking away opportunity. It's the obvious recognition that learning from mistakes needs room to fail safely, and that a world with no possible failure is also a world with no possible learning.

If you lead others, there's a harder version of all this, and it's simply that their mistakes are actually yours. You hired them, or trained them, or built the room they work in. The system is yours to own. The work is yours to be accountable for. The good ones don't punish — they make sure the learning happens, for all involved – including themselves – and that the system of work gets a little better each time. It's tiring, unglamorous work, and rarely makes it to a board's recognition deck, but it's most of what the job actually is.

Because the enemy was never the mistake itself. It's the wasted potential of not learning from it. Studied, a mistake becomes refinement; refinement repeated becomes something close to mastery. That's a very positive way to look at it, and that's what I wrote down during the talk on a small piece of paper.

Failure is not about being fast. Failure is also not an identity. Failure cannot be avoided if meaningful work is being done. Failure needs a safe space to land. And failure is actually information – if we choose to look.