A practical stress-test for finding out what the business actually depends on you for, before a buyer or successor does.
Key takeaways
Ask most owners whether the business could run without them for a month, and they'll usually say some version of "mostly, yes". Actually leaving for a month tends to produce a much more specific and useful answer, because it replaces a guess with a list of exactly what happened while they were gone.
This is worth doing deliberately, not as an accident of a vacation cut short by phone calls. The value comes from genuinely stepping back and observing rather than staying reachable in practice.
On return, what happened during the absence usually falls into three groups: decisions that stalled completely because nobody else was authorised or confident enough to make them, decisions someone else made adequately, and decisions someone else made in a way the owner wouldn't have. Each category points to a different fix.
Stalled decisions point to missing authority or a missing process. Decisions made differently point to something that was never actually explained, just assumed. Only the third category, things that genuinely went wrong, points to a training or judgment gap.
Whatever stalled or went sideways during the month away is a far better guide to what to document or systemise next than any guess made in advance. It replaces "what do I think matters" with "what actually mattered, once I wasn't there to catch it".
This is also a much cheaper way to find gaps than discovering them during an actual health scare, a real sale process, or a family emergency, when there's no controlled way to step back in if something goes wrong.
This only works if the team knows it's happening and understands the point is to find gaps in the business, not to catch anyone out. Run as a surprise test, it breeds resentment and produces worse information, because people become guarded rather than making their normal judgment calls.
Framed honestly, as wanting to find out what actually needs fixing before it matters more, it tends to get genuine cooperation, and often surfaces problems the team has been quietly working around for years.
A 30-minute call is enough to tell you whether AI pays for itself here.
