Stop talking about performance. Start measuring it.

I’ve led operational transformations across multiple markets. The difference between companies that scale and those that stall isn’t talent, budget, or strategy.

It’s the willingness to work from one set of numbers.

It sounds simple. It’s actually a cultural battlefield. Here is the “no-fluff” guide to making metrics work:

If you can’t measure it, don’t prioritise it

We all have initiatives that feel important. Culture improvements. “Better collaboration.” Process optimisation that’s hard to pin down.

The uncomfortable truth: if you can’t measure progress, you can’t manage it. You’ll end up in quarterly reviews where everyone agrees things are “moving in the right direction” without anyone being able to say what actually changed.

This doesn’t mean everything needs a KPI. But the things you claim are priorities? Those do.

If you can’t find even a creative way to measure it, then drop it.

Data quality won’t improve until you start using it

This is the one that trips up most organisations.

Stop saying, “We’ll use the reports once the data is clean.” That day will never come.

Data quality only improves when people have a real reason to care about it.

So put the numbers on a screen in the Monday meeting even though you know they are wrong. Nothing motivates people to keep their data clean more than knowing that everyone else will see it.

If the data is actually created by teams elsewhere, bring the numbers to them. Teams that enter data but never see the output is a recipe for bad data.

P.s. If you can, it is a good idea to do an initial cleanup sprint. Book the right people for one or two days, or bring in temporary help, and fix the obvious issues. Just don’t wait for everything to be perfect.

Your data isn’t locked. You just haven’t fetched it yet

“We’d love to track that, but the data x is not in that system.”

Five years ago, it was sometimes true. Today, it almost never is.

Modern tools, especially AI-powered ones, can pull data from nearly anywhere. ERP systems, CRMs, spreadsheets that someone updates manually, even PDFs and emails. The technical barriers that used to exist have largely disappeared.

No one on your team knows how to connect the systems? Find someone on Upwork or Fiverr who can. It’ll cost you a few hundred euros and save you months of operating blind.

The data exists. Go get it.

Show data when people can actually act on it

Timing matters more than most people realise.

Daily metrics belong on a screen in the office, somewhere people walk past throughout the day. Not in a dashboard they have to log into. (Note: Screens in the office are a surprisingly effective way to get people focused on the numbers)

Weekly metrics belong in the Monday meeting, not in a report that gets emailed Friday afternoon and forgotten by the time anyone could act on it.

Monthly metrics should be reviewed right around month end. Is finance only closing the books on the 10th business day? Then find proxy numbers to review.

The principle is simple: numbers only drive behaviour when the feedback loop is tight enough for people to do something about what they see.

Don’t overload. Phase it in

When you finally commit to data-driven management, the temptation is to do it all at once. Fifty KPIs. Dashboards for everything. Complete visibility from day one.

This backfires almost every time.

People can only absorb so much change. When you put too many numbers in front of them, they tune out. Everything becomes noise, and nothing drives behaviour.

Better approach: roll it out in stages over two to three months.

Start with the two or three metrics that matter most. Get people comfortable reviewing them, discussing them, acting on them. Build the habit.

The real barrier isn’t technical

The tools exist. The data is accessible. The methods are well understood.

What separates organisations that succeed with metrics from those that don’t is simpler and harder: the willingness to commit.

Commit to one source of truth. Commit to reviewing it consistently. Commit to holding people accountable to numbers they might not like. Commit to fixing the data instead of dismissing it.

One set of numbers, reviewed regularly, with consequences attached. This beats sophisticated dashboards that nobody opens.