Track rightsizing progress

Every rightsizing plan makes claims about the future: this capacity will free up, that work will redistribute cleanly, coverage will hold, these people will transition by this date. Then the plan is approved, the organisation exhales, and the predictions go unexamined until the quarterly review, by which point every variance is three months old and expensive to correct. The plan was a forecast; nobody stayed to watch the weather.

The most dangerous assumption in any such plan is that work disappears with the role. It usually does not. It redistributes, silently, onto the nearest capable team, and for a quarter or two that team absorbs it invisibly through overtime and goodwill. The plan looks on track precisely because people are quietly paying its costs. By the time the strain surfaces in resignations, you are losing exactly the people the plan meant to keep.

Teams do not announce they are drowning. The data does, if you are watching it.

The humane and the commercial answer turn out to be the same one: watch the leading indicators. Overtime creep, ticket backlogs, cycle-time slips on affected processes, all of it visible in systems you already run, all of it surfacing weeks before it shows up in exit interviews. A steering meeting armed with that data makes corrections while they are still cheap nudges: resequence a milestone, backfill a transition, slow one workstream. A steering meeting armed with anecdotes argues about whose impression is right.

This programme keeps the plan honest: every milestone tracked in the graph as it lands, every freed hour traced to where it went, and the strain signals watched so the plan gets corrected by evidence rather than by attrition.