In short: Reasons leak from organisations all the time. At handovers they pour out, all at once, on a date everyone knows in advance. That makes the handover the cheapest, most obvious place to start a judgment record, and the most expensive one to skip.
Key points
- A standard handover transfers what is already written down and loses what is not, which is the part that mattered.
- The pack covers extracted decisions, open threads, relationships, the graveyard of abandoned attempts, and a 90-day question log.
- Extract decisions in a 60 to 90 minute structured interview, asking what the replacement should not do.
- Start four weeks before departure and schedule a 30-day follow-up with the outgoing person.
- Repeat failures, abandoned approaches tried again without new evidence, should be zero.
A programme lead rotates. A committee's season ends and a new panel starts cold. The first commercial hire inherits a CRM and a folder of half-finished documents. An agency's work comes in-house. An advisor's engagement closes. Each time, the reasons behind dozens of live decisions leave with the person, usually within a two-week notice period spent mostly on other things.
The standard handover makes this worse, not better. It is a document dump: the files, the logins, a status list. It transfers what is already written down and loses what is not, which is the part that mattered.
The five parts of a handover pack
1. Decisions, extracted. The outgoing person's ten to twenty most significant recent calls, each with the situation, the choice, what was rejected, why, and what happened. Done as a structured interview, not a form, because nobody writes these down unprompted. Sixty to ninety minutes is usually enough. One question reliably produces the most valuable answer: "What would you tell your replacement not to do?"
2. Open threads. Every decision in flight, with its current state, the next move, the date it matters by, and what would change it.
3. Relationships. Who matters and why. What was promised to them, and by whom. What they are sensitive about. This is the part that exists only in the outgoing person's head, and it is the part whose loss is noticed last and costs most.
4. The graveyard. What was tried and abandoned, why, and what would make it worth trying again. The incoming person's first instinct will often be to try one of these. The graveyard tells them whether that is a good idea.
5. A question log for the first 90 days. The incoming person asks the record before asking around. Every question the record cannot answer is logged, which shows exactly what the handover missed, and is filled while the outgoing person can still be reached.
Timing
Start four weeks before departure, not in the last week. The extraction interview should happen while the person is still engaged with the work, not mentally gone. Schedule a 30-day follow-up with the outgoing person, by agreement, to answer the accumulated questions from the log.
Why it is the best place to start
Handovers are timed. You know they are coming. They have an obvious owner, the incoming person, who is highly motivated to get the reasoning. And they produce immediate, visible value: the new person makes fewer avoidable mistakes in their first quarter.
For an organisation with no judgment record at all, the handover is the easiest point at which to begin one. Each handover seeds a decision record, a graveyard and a relationship map for that role. After a few, the organisation has the start of a record without ever having launched a programme.
How to tell if it worked
- Time to first independent decision for the incoming person.
- Share of "why did we…" questions answered by the record rather than by finding a human.
- Repeat failures: abandoned approaches tried again without new evidence. This should be zero.
- Questions still unanswered at day 90.
In practice
- For the next three departures in key roles, run the five-part pack starting four weeks out.
- Use the "what would you tell your replacement not to do?" question in every extraction interview.
- Give the incoming person the question log and review it at 30 and 90 days.
- Keep the packs. Together they become the first version of the organisation's judgment record.
What would prove this wrong. Incoming people with a five-part pack ramping no faster, and repeating no fewer abandoned approaches, than those with a conventional handover.
Common questions
What should a good handover document include?
Five parts: the outgoing person's ten to twenty most significant recent decisions with reasons, every open thread, key relationships and what was promised, a graveyard of what was tried and abandoned, and a question log for the first 90 days.
When should a handover start?
Start four weeks before departure, not in the last week, while the person is still engaged with the work. Schedule a 30-day follow-up with the outgoing person to answer questions logged since.
What is the best question to ask someone leaving a role?
"What would you tell your replacement not to do?" reliably produces the most valuable answer in the extraction interview.
How do you know if a handover worked?
Track time to first independent decision for the incoming person, the share of 'why did we' questions answered by the record, repeat failures, and questions still unanswered at day 90.