Main idea

The committed column of a roadmap is not sized by management's ambition, by sales pressure, or by the number of interesting topics. It is sized by what the team can run well at the same time: two major topics, three if they are smaller.

Displaying eight does not make the company more ambitious. It produces an object that lies about itself: every topic in it bears the mark of commitment, none has the reality of it. The teams know half of them will not move, the sales force relays commitments that are none, and the roadmap stops being legible at the precise moment it was asked to clarify.

The same thing shows in its length: a roadmap that no longer fits on one page has stopped sorting. It has redeposited, in the form of three columns, the stockpile it was supposed to reduce.

Why it matters

It gives an observable sign that can be checked without knowing the product: compare the number of committed topics to the number a team of that size can carry. A clear gap is enough to know no arbitration took place.

It also names the exact cost of overload, which is not delay. It is the object's loss of credibility: once people know a topic displayed as committed may be nothing at all, no displayed topic commits anything any more.

Nuances and limits

Real capacity is hard to establish and open to debate — a team rarely overestimates it in good faith, an executive rarely underestimates it. The rule says what to calibrate the column against, not how to measure the quantity.

And a column that is too short has its own flaw: it leaves capacity unused, which the team will fill with topics that will never have been arbitrated at all.

Open questions

• How do you establish a team's capacity other than after the fact, when topic sizes vary by a factor of five?