You Will Never Have Enough Information. Decide Anyway.

I once held a decision open for seven weeks because I wanted one more number. When the number finally turned up it was a few percent off the estimate we'd been carrying all along, and it changed absolutely nothing about the answer. The seven weeks changed quite a lot.

Here's what makes deferral so seductive. If you decide and you're wrong, that's yours. Your name is on it, people remember, and someone will bring it up in a steering group nine months later. If you wait and the programme slips, the slip belongs to nobody. It gets written up as "awaiting clarification" and everyone nods, because the alternative is admitting that a person chose to spend seven weeks.

So deferral feels safe. It isn't safe, it's just quiet.

While you're waiting, your team is still working. They're building on whatever assumption seemed most likely, because they have to build on something. Every week you stay undecided, more work gets committed to an answer you haven't given, and the eventual decision gets more expensive to implement whichever way it goes. You didn't avoid the risk. You let it accumulate somewhere you can't see it.

The hardest version of this to spot is when waiting genuinely is correct. Sometimes the information arriving on Thursday really does change the answer. The difference between a good leader and a stalling one is that the good one can tell you, out loud, what she's waiting for and what she'll do when it arrives.

This is the single most useful question I know, and it takes about eleven seconds.

Before you commission another analysis, ask: what would this have to say for us to choose the other option? Write the answer down. If you can't write one, or if every plausible answer leads to the same choice, then the analysis isn't information. It's comfort, and you're buying it with calendar time you don't have.

The same question works on other people. When someone tells you they can't commit until they've seen the numbers, ask them which number and which way. Half the time you'll discover they don't want more data, they want more cover. That's a completely human thing to want and it's worth naming kindly, because the fix is different. They don't need a spreadsheet. They need you to say clearly that the call is yours and their name isn't on it.

Involve the people who'll have to live with the outcome. Consult the ones who know the thing you don't. Then decide. Consensus is not a requirement, and the search for it is the most respectable way in our industry to avoid doing your job.

A car rental company, 260 branches, a full network refresh with a hard deadline set by a circuit contract expiring at the end of the financial year. We needed to place the hardware order. Lead time on the routers was eleven weeks and getting worse.

The question was whether to order 260 units of the standard build now, or wait for a physical survey of every branch to confirm power, rack space and cabling before committing. The estate records were about six years old and everyone agreed they were unreliable.

I waited for the surveys. It felt like rigour. It was the option that nobody could criticise in a meeting, which should have been my clue.

The surveys took seven weeks because they needed branch managers available outside counter hours, and a good slice of the estate was airport desks that only go quiet after the last flight lands. They came back showing that 9 branches out of 260 needed a different enclosure. Nine. We could have ordered 260 standard units on day one and handled 9 exceptions with a phone call and a small second order, which is exactly what we ended up doing anyway, just eleven weeks later on the delivery clock.

What I should have asked myself: what would the survey have to find for us to order something other than the standard build? The honest answer was "more than a third of the estate", and nobody in the room believed that was likely. That answer was available on day one. It was free. I just didn't ask.

The bit that stings is that the survey was still the right piece of work. It just wasn't a blocker. Running it in parallel with the order would have cost nothing and told us the same thing.

Memory is a liar, and it's a liar with an agenda. Six months after a decision goes badly, you will genuinely remember yourself having had doubts you never had. Six months after one goes well, you'll remember being more certain than you were. Both are corrosive. The first makes you timid, the second makes you reckless.

So write the decision down when you make it. Not a document, not a template with a version number. Five lines in whatever your team already reads:

What we decided. What we knew at the time. What we assumed and couldn't verify. What would make us revisit this. Who owns it.

That last line matters more than the rest put together. A decision without a named owner isn't a decision, it's a mood in a room.

The "what would make us revisit" line is the one people skip and then regret. It converts a decision from a thing you defend into a thing you monitor. When the trigger fires, changing course isn't a humiliating reversal, it's the plan working. I've watched leaders burn real credibility defending a call for months after the ground shifted, purely because they'd never written down what would make them change it, and reversing felt like weakness rather than what it actually was.

And when it does go wrong, you own it out loud. Not "the information we had at the time was incomplete", which is the sentence people use when they want the passive voice to take the hit. Try "I chose this, here's what I got wrong, here's what we're doing now". It costs you about four seconds of discomfort and buys you years of people believing what you tell them.

Open whatever list you keep, and find the decision that's been open longest. Not the biggest one, the oldest one. There'll be one that's been quietly rolling from week to week since roughly June.

Write two sentences and send them to the people involved today. The first says what you'd choose if you had to choose right now. The second says what would have to be true for you to choose differently, and by when you need to know it.

Then put a date on it, this week if you possibly can, and put your own name next to that date.