Every company I join has a roadmap. Most have several: one in the board deck, one in the project tool, one in the CEO’s head. None of them agree, and all of them are too long.

Why one page

A page is a constraint that forces decisions. When the roadmap is a 40-row spreadsheet, nothing has to lose. When it is one page, every item has to beat the others for space, and the team can hold the whole plan in their heads.

At Halden we went from 41 items to 6. Release frequency tripled¹, mostly because people stopped context-switching between projects that were never going to ship.

The template

Three columns: Now, Next and Later. Now has at most three items, each with a customer problem, a metric and an owner. Next has at most five. Later is a list of names, nothing more. Below it, one line: what we are deliberately not doing this quarter.

Not doing this quarter: one line, written down.

The ‘not doing’ line is the most important line on the page.

From my Halden notes, 2025

How to cut

Put every item in a sheet. Ask each owner for the customer problem and the evidence. Anything without evidence goes to Later. Anything that cannot name a metric goes to Later. You will be left with fewer items than you expected, and that is the point.

Try this: Before your next planning meeting, ask every owner for the customer problem and one piece of evidence. Count how many items survive. In my experience it is about one in seven.

Then show it to sales before you show it to the board. I learned that one the hard way.²

Notes

Halden case study, weeks 2 to 13. Measured as production releases per month, from 0.7 to 2.2.

Two of the cut items had been promised to named prospects. See ‘What I would do differently’ in the Halden case study.