Handover should be a data structure, not a meeting

A handover meeting creates alignment for an hour. A governed handover record creates it for the life of the engagement.

The difference is not effort, seniority or goodwill. It is which specific fields travel out of the sale and into delivery, and whether they travel as data a system can enforce or as prose a person has to read, remember and interpret.

Why the meeting feels sufficient

The handover meeting is the most reassuring hour in a professional services firm. Everyone who matters is in the room. The proposal is on screen. Questions get asked and answered. Nobody leaves confused.

That is exactly the problem. The meeting produces shared understanding in a set of heads, and heads are a storage medium with no version control, no access log and high staff turnover.

Six weeks later the delivery lead has changed, the client has asked for something adjacent to scope, and the commercial basis for saying yes or no exists as a recollection of a conversation. The alignment was real. It just had a half-life.

The mistaken diagnosis

Most firms read this as a documentation problem, and respond by making the handover pack longer. A better template, a mandatory checklist, a fuller set of notes attached to the project folder.

A longer document does not fix it, because a document is still something a human must consult voluntarily. The failure mode is not that the information was never written down. It is that nothing downstream is obliged to obey it.

A billing rule in a PDF cannot stop an unbillable hour being recorded. An assumption in paragraph four cannot flag a change of scope. A named approver in a table cannot refuse an approval. Prose describes the commercial position; it does not hold it.

What actually has to travel

This is the whole post. A handover is a data structure when six things exist as fields on the engagement record rather than as sentences about it.

Scope, with its assumptions attached. Not “website redesign” but the deliverables, and against each one the assumptions that made the estimate honest: how many review rounds, which client resource was expected to be available, what counts as complete. The assumption belongs to the deliverable it prices, not to a paragraph on page two.

Budget at three levels. Contract, milestone and task. A single contract value tells you nothing until the job is over. Budget held at task level is what lets a project manager see the overrun in week three rather than at close.

Billing rules as data. Which time is billable, at which rate, under which conditions, with which cap. Rate exceptions and discounts agreed during the negotiation belong here, in the field the invoice reads from — not only in the contract, where they are frequently agreed and then never applied.

Acceptance criteria. The test a deliverable has to pass to be complete, recorded as a state that gets set, not a clause that gets quoted during a dispute.

Named approvers, on both sides. Who can approve time, who can approve a variation, who can accept a milestone, and who at the client can commit to it. “The client signed off” is not a fact until it names a person.

Change-control thresholds. The point at which a change stops being absorbed and becomes a priced variation, agreed before delivery starts rather than negotiated at the moment it is needed. A threshold set in advance is a policy. The same threshold set in the moment is a concession.

The test

There is a simple way to tell which kind of handover you run. Take a live engagement and ask the system, rather than a person, three questions: what is this project allowed to bill for, who can approve a variation on it, and what did we assume when we priced it.

If the answers come back as records, the handover was a data structure. If they come back as “ask the account lead”, it was a meeting.

What makes this workable in practice is that the engagement delivery is built against the record that was quoted, so the plan inherits the commercial basis instead of restating it — and every time and cost entry made afterwards inherits the rules that decide whether it can be billed. There is more on where those terms get set on our scoping and quoting page, and on what happens to them once work starts on our delivery and time capture page.

What this costs

Structuring the handover moves work forward into the sale, where it is least welcome and hardest to enforce. Somebody has to record assumptions, thresholds and approvers at the point where the commercial instinct is to close and move on. That cost is real, it lands on senior people, and no software persuades them to pay it.

It also slows the first week. A firm that wins on responsiveness — team on site the Monday after signature — will feel a gated handover as friction, and for a two-week engagement the overhead can genuinely exceed the leakage it prevents. The discipline earns its keep on engagements long enough or complex enough to drift.

And it is not free of judgement. Fields make the position explicit, which means disagreements that used to surface quietly at invoicing now surface loudly at kick-off. That is the mechanism working, but it does not feel like an improvement in the first cycle.

Structure, not ceremony

Keep the meeting. It is useful for context, relationship and the things that genuinely do not reduce to fields — client politics, the personalities on the other side, what the sponsor is actually being measured on.

Just stop asking it to carry the commercial position. A meeting is how people understand an engagement. A record is how the engagement survives the people. If you want to see what one governed record looks like across the whole chain, our invoicing and WIP page follows a single entry from timesheet to invoice line.