Operations problems are rarely about how hard the work is. They are about how often the work gets handed from one person or system to another. Each handoff is a potential stall, a chance for context to drop, and a place where ownership goes soft. The five-handoff rule is a forcing function for designing them out.
The math explains why the rule bites. Give every handoff a modest 90% chance of going cleanly (context intact, picked up promptly, nothing repeated) and a five-handoff workflow completes cleanly about 59% of the time; at eight handoffs it is under 43%. Nobody in the chain sees the compounding, because each individual handoff really is fine most of the time. The workflow, as a whole, is a coin flip, and everyone blames bad luck.
The rule
For any end-to-end workflow, count the handoffs. If there are more than five between the trigger and the outcome, the workflow will leak, regardless of how good your people are. Design the process to reduce the handoff count, not just to make each handoff faster.
Count honestly, and the number is always higher than anyone guesses. A handoff is every transfer of responsibility, including the invisible ones: person to system (filed into a queue), system to person (waiting on an approval), team to team, and person to themselves-next-week (parked in a someday list). The onboarding flow the sales team describes as "we hand it to delivery" is usually six or seven transfers once the forms, approvals, and tool-switches are counted. Walking one real item and tallying is a twenty-minute exercise that changes the conversation permanently.
How handoffs leak
The handing party assumes the receiver knows context they do not. The receiver waits for clarification, which arrives slowly. The work sits. Status is unclear. Eventually the customer or the next team chases. Each leak is small; the cumulative tax is enormous.
The deepest leak is ownership diffusion. Between the moment one person lets go and the next truly takes hold, the work belongs to nobody, and those ownerless gaps are where days disappear: not in anyone's queue, just between queues. Ask who owns an item mid-handoff and the honest answer is a shrug in both directions. Customers can feel these seams from the outside, which is why processes with many handoffs read as unprofessional even when every individual is excellent.
What to do about it
Compress steps that do not need to be separate roles. Automate handoffs that follow a clear rule. Add a single source of truth so handoffs do not require restating context. The goal is not heroic operators; it is a workflow that does not need heroes.
The order of operations matters: eliminate before automating, because automating an unnecessary handoff just makes the waste faster. For the handoffs that survive, make each one a designed object with three properties: a standard payload (what context always travels with the work, so nothing needs re-asking), an explicit acceptance (the receiver confirms taking ownership; silence does not count), and a timestamp, so aging handoffs surface on their own instead of waiting for a chase. And treat the transitions as steps with owners, not gaps between steps: the single change of naming an owner for every seam converts the ownerless days into accountable hours.