Flux8 All articles
Opinion

Ask Where They Guessed: The Project Review That Fits in One Sentence

Flux8
Ask Where They Guessed: The Project Review That Fits in One Sentence

Photo by Photo by Juno Jo on Unsplash on Unsplash

There is a particular kind of silence that follows the question "so, what went well?" in a UK project review. It's not hostility. It's politeness, plus a mild panic about saying something that could be read as criticism of a colleague who is, unfortunately, in the room.

So people say the workshop was useful. They say communication was good, considering. They say the team pulled together. Everyone nods. The meeting ends. Three weeks later the same project's successor hits exactly the same pothole, and nobody can quite explain why the lesson didn't stick.

The lesson didn't stick because it was never found. It's not that people are hiding it. It's that the interesting failure mode on most projects isn't a blow-up. It's an accumulation of small uncertainties that everyone absorbed privately and worked around silently.

One Question, Asked Properly

Try this instead. Get the same people in the same room and ask a single thing:

"Where did you have to guess?"

That's the whole review. Not what went well, not what didn't, not a RAG status retrospective. Where did you have to guess.

The question works because it's guilt-free. Nobody has to accuse anyone. Nobody has to admit to a mistake — guessing is what sensible people do when information is missing, and it is a completely reasonable response to a badly documented process. You're not asking who dropped the ball. You're asking where the ball's flight path was unclear.

Ask it in the right voice, though. Said briskly at the end of a ninety-minute meeting, it lands as an audit. Said at the start, with a pen out and a genuine follow-up question ready, it lands as an invitation.

What Comes Back

The answers are almost always smaller than people expect, and more useful.

A designer guesses which of three brand versions is current. An account manager guesses whether the client actually signed off the copy or just didn't object. Someone guesses whether the new starter has access to the reporting suite yet. A developer guesses whether that email from the client counted as approval. Two people guess, differently, at what the deadline actually was.

None of these are dramatic. All of them cost time — usually in the form of a rework, a delay, or a slightly awkward month-three conversation. And each one is a candidate for a fix that takes less than an hour: a named owner, a folder, a one-line rule, a template, a decision log nobody has to be trained on.

Keeping It From Turning Into a Trial

A review that invites people to describe uncertainty can curdle into blame very quickly if you let it. Three guardrails help.

Collect first, discuss second. Everybody writes their guesses down before anyone speaks, and you read them out without attribution. If the first answer discussed becomes the loudest one, you've lost the room.

Record the guess, not the guesser. The output is a list of gaps in the process, not a list of people who didn't ask enough questions. Write it that way.

Never ask "why didn't you check?" The honest answer is usually "because checking would have taken two days and I had an afternoon." That's a workflow finding dressed up as a personal one.

Ranking Them Without Overthinking It

You'll usually get somewhere between eight and twenty guesses from a small project. Don't score them out of ten. Just sort into three piles:

The top three recurring or expensive items become actions with names and dates. Everything else goes in a list that gets read at the next review. That's the whole ceremony. It's genuinely that light.

Write It Where People Will Find It

The most common failure of any review is that its conclusions live in a document nobody opens. Guesses are different. If you've discovered that six people didn't know which brand version was live, the fix isn't a lesson-learned paragraph — it's a folder rename, a pinned note, or a single line in the project template.

Make the fix land in the place the work actually happens. Not the intranet. Not the retrospectives folder that everyone is vaguely aware exists. The place a person would look at 4:50pm on a Friday when they need to know whether to send the final version.

Why It's Worth the Ten Minutes

In 2026, most UK teams have more process documentation than they've ever had, and people are guessing more than ever — because the documentation is spread across four tools, two of which replaced something else six months ago. Volume isn't the problem. Proximity is. The answer has to be next to the moment of work.

Asking where people guessed is a cheap way to find the exact spots where proximity fails. It takes ten minutes, it doesn't require a facilitator, and it produces a list of small, specific things you can actually change.

And there's a secondary benefit that never gets mentioned: it tells your team that admitting uncertainty is a normal, useful thing to do. That's not nothing. That's most of the culture work, done in a sentence.

All Articles

Related Articles

The Process You Killed Is Still in Someone's Inbox

The Process You Killed Is Still in Someone's Inbox

The Bot Has No Boss: Why AI Agents Need a Named Owner Before They Get More Power

The Bot Has No Boss: Why AI Agents Need a Named Owner Before They Get More Power

Who Changed the Process? Your Workflow Needs a Change Log, Not Another Dashboard

Who Changed the Process? Your Workflow Needs a Change Log, Not Another Dashboard