Start with roles, not names
Succession work goes wrong early when it begins as a list of people. Begin instead with the roles the organisation cannot afford to leave open or fill badly — the ones where a poor appointment would cost time, customers, safety or credibility.
Keep that list short enough to act on. A critical-role list that includes every senior title is a directory, not a plan. Once the list is honest, ask a single question of each role: if the holder left this quarter, what would we actually do?
Separate coverage from readiness
Coverage answers whether anyone is named against a role. Readiness answers whether that person could hold it now, soon, or later. Many organisations report strong coverage while carrying almost no ready-now capability, because the same few names appear against several roles.
Count each successor once. Where one person covers three roles, you have one successor and two exposures. That single discipline changes the shape of most succession maps.
Make readiness evidence-based
Readiness judgements formed only in the room tend to reward visibility. Anchor them in something more durable: delivery against stretch scope, behaviour under pressure, feedback from people who work alongside the person rather than above them, and structured assessment where the stakes justify it.
Assessment data is an input to a conversation, not a verdict. Used well, it makes calibration faster and less political, because the discussion moves from impressions to specifics.
Close the loop between review and development
The most common failure is not the review itself but the silence afterwards. If the talent review does not produce owned development moves — a scope change, a project with real consequences, coaching, an exposure to the board — then next year's review repeats this year's conclusions.
Give every ready-soon successor one development commitment with a name and a date attached. Review the commitments, not just the grids.
Where intelligent tools help — and where they do not
Technology is good at the parts humans do badly at scale: pulling movement, tenure and performance signals together, flagging concentration risk, surfacing people the process usually overlooks, and keeping the picture current between formal cycles.
It is not good at judging whether someone will hold their nerve in a difficult year. That judgement stays human, informed by better evidence and made in a room where people are willing to disagree.
What to take away
- Define critical roles by consequence of failure, not by seniority.
- Count each successor once — shared names hide real exposure.
- Distinguish ready now, ready soon and ready later, and say why.
- Attach one owned development commitment to every ready-soon successor.
- Use tools to keep the picture current; keep the judgement human.