You Have Dashboards. You Still Don't Know How Last Week Went.

Every tool you use ships with analytics, so most companies now have more dashboards than decisions. The result is a strange kind of blindness: an enormous amount of data and no shared answer to the question of whether last week was good.

Track what you control, not what is easy

Most dashboards default to what the tool happens to measure. Page views, tickets closed, emails sent. These are easy to collect and nearly impossible to act on, because no single person can change them directly.

A useful number has an owner who can move it this week by doing something differently. If nobody can, it is context, not a metric — interesting at the board meeting, useless on a Monday.

Build the tree downwards

Start with the one company number that matters this year. Break it into its two or three drivers. Break each driver into the thing a specific team controls. Stop when you reach something a named person can change inside a week.

You will end up with somewhere between five and nine numbers in total. That is the right amount. Thirty numbers is not thoroughness; it is an unwillingness to decide what matters.

The sheet beats the dashboard

This will sound backwards: for most companies under fifty people, a spreadsheet somebody fills in on Sunday beats an automated dashboard. Not because automation is bad, but because the act of entering the number is the act of noticing it.

Automate it later, once everybody already knows what the numbers mean and why they moved. Automating too early produces a chart nobody can interpret and nobody owns.

Twenty minutes on a Monday

The review matters more than the report. Twenty minutes, same time every week, same shape every time: what moved, why, what we are changing. Owners speak about their own numbers.

Most importantly, the meeting ends with decisions, not observations. A review that produces no change in behaviour is a reading group.

Then a monthly pack that takes an hour

If your monthly reporting takes a week to assemble, nobody will do it honestly by month four. Build it from the weekly sheet so the month is just an aggregation, and keep the commentary to what changed and what you did about it.