← All posts
Heretical Take   May 11, 2026 · 5 min read

Why OKR heatmaps are thermometers, not doctors

Generated illustration for the post 'Why OKR heatmaps are thermometers, not doctors'

Open any modern OKR tool — Cascade, Lattice, Quantive, the one your competitor just demoed — and the first screen will be a heatmap. Green, amber, red, neatly arranged by team or objective. Executive-friendly. Quarterly-read-out-ready. Designed to photograph well in a sales deck.

It is, almost without exception, useless when the moment arrives that it was supposed to matter for.

What a heatmap actually tells you

That something is off. That's the entire payload. A red cell says: this KR is behind target. It does not tell you which work item was supposed to move it, whether anyone is actively working on that item this week, whether the right person owns it, or what you should start, stop, reassign, or kill before Friday.

You can stare at a beautifully designed red square for fifteen minutes and learn nothing actionable. The heatmap is a temperature reading. The reason the patient has a temperature lives one layer down — in the actual work that should have been moving the number.

I've watched executive teams do this dance every quarter. Someone clicks into the red cell. The owner gives a verbal update. The room nods. Nothing on the screen changes, because the screen doesn't know about the work. It only knows about the score.

Why every vendor keeps shipping them anyway

Because heatmaps demo well. They look like clarity. They photograph nicely for a board pack. And — this is the uncomfortable part — they let an executive feel they have visibility without anyone having to do the harder work of connecting the goal to the work meant to move it.

Is the heatmap clarity, or is it a way of feeling informed without being informed? In most companies I've sat in, it's the second one. The heatmap becomes a ceremonial object: presented, acknowledged, archived, repeated next quarter.

A red heatmap is the start of a question. Most tools end there. That's the whole business model of a generation of OKR software: end where the actual operating problem begins.

What "the doctor" actually looks like

A real strategy execution layer, when a KR is at risk, can immediately show you four things on one surface:

That isn't a dashboard. That's an operating layer. It's the thing a Chief of Staff currently rebuilds by hand every Sunday night because no tool does it for them — six tabs, two Slack channels, three spreadsheets, a calendar, and a lot of asking around.

The capacity question hiding behind every red cell

The other thing the heatmap will never tell you: whether the red is caused by a wrong bet, a missing owner, or a capacity collision. Those three failure modes look identical on a dashboard and require completely different responses. A wrong bet needs a strategy conversation. A missing owner needs a naming conversation. A capacity collision needs a deallocation conversation — and the only honest version of that conversation requires knowing which other goal you're going to deprioritize to free the person.

None of that lives in the colour. All of it lives in the graph underneath. If your tool only renders the colour, you'll keep having the same meeting every quarter and reaching the same conclusion: we need better discipline. You don't need better discipline. You need a system that shows you the work, the owner, and the capacity at the moment you're looking at the red.

The Vindaris view

A heatmap is a fine summary artefact for a board slide. It is not, and has never been, a strategy execution system. If your tool stops at colour-coding, you don't have execution software. You have a thermometer with a nice login page — and a Chief of Staff still doing the doctor's job on Sunday night.

The category will keep building prettier thermometers. The actual unlock is one layer down: goals, work, owners, and capacity in the same graph, queryable in real time, traceable from the red cell to the engineer whose Tuesday is supposed to be about it. That's the meeting nobody has yet because the data nobody has yet.

The next time your leadership team stares at a red square together, ask the room one question: what specifically would have to change between now and Friday for this cell to turn amber, and who is going to make that change? If nobody can answer in plain language with a name and a date, the heatmap has told you exactly nothing — and the work of turning the colour back lives somewhere the tool can't see.