Revenue leakage is the cost of re-explaining the work

Most revenue leakage is not caused by bad work. It is caused by the same job being told three different ways.

What was sold. What was delivered. What can be billed. In a firm without a single governed record, those are three separate accounts of one engagement, each produced by a different function, each internally consistent, and each requiring translation before it can be reconciled with the others.

The translation is where the money goes.

The mistaken diagnosis

Leakage is usually read as underperformance. Someone over-serviced the client, someone estimated badly, someone did not chase the variation. The response is a performance response: tighter budget discipline, a margin conversation with delivery leads, a reminder about scope.

That reading survives because it is occasionally true. But it does not explain the pattern, which is that the leakage is remarkably consistent across engagements, across delivery leads, and across quarters — including on projects everybody agrees went well.

Consistent leakage on good work is a system property, not a people property. The work was fine. The story about the work stopped carrying.

Where the two breakdowns happen

There are two specific boundaries, and almost all of it happens at one of them.

Sales to delivery. What was sold is not handed over as governed, usable input. Scope, assumptions, rates and commercial conditions arrive as a proposal document, an email thread and a verbal briefing. Delivery therefore begins by interpreting. Every interpretation is a small unpriced decision about what the client is owed.

Delivery to finance. What was delivered is not handed over as billable evidence. Time, costs, changes and approvals sit across several tools in several states of completeness. Finance therefore begins by chasing and reconstructing, weeks after the fact, with less context than anyone who was there.

Neither boundary fails dramatically. Both fail by requiring the job to be re-narrated by someone who was not present for the previous telling.

Where the concession is actually made

This is the part that rarely gets attributed properly.

The loss is not booked at the boundary where it was created. It is booked later, at the point where re-explaining becomes uncomfortable — usually in front of a client, usually under time pressure, usually by whoever is holding the relationship.

It looks like this. A line item softened because nobody can produce the approval for it. A variation billed at the original rate because the rate change was agreed on a call. Hours written off because the time was entered against the wrong engagement and correcting it properly would take longer than the hours are worth. A credit note issued to close a conversation rather than to correct an error.

Each of those is a judgement call, and each is defensible on its own terms. Collectively they are the leakage, and they are invisible in reporting because they are recorded as commercial decisions rather than as consequences of a hand-over that did not carry.

That is also why the leakage does not appear as a variance anyone can act on. It appears as margin that was always going to be a bit lower than the quote — which is how leakage becomes normalised. The wider version of that argument is in WIP: register or argument.

The mechanism

Removing translation means the same record has to survive both boundaries.

Concretely: the scope carries its assumptions as fields rather than as prose, so delivery inherits the conditions the price depended on. Budget sits at contract, milestone and task level so a burn question has an answer at the level it is asked. Billing rules are data the system applies, not a clause someone reads. Named approvers and change-control thresholds are attributes of the engagement, so a variation has a defined path before anyone needs one. Where those rules get set is covered on our scoping and quoting page, and what happens to them downstream is on the invoicing and WIP page.

When that holds, delivery does not interpret and finance does not reconstruct. The invoice line is assembled from records that already exist rather than from an account of what happened.

What this costs

Two honest caveats.

Not all re-explaining is waste. A client relationship in which nobody ever restates the commercial position is not automatically healthy — restating scope at a steering meeting is how expectations stay calibrated, and a firm that treats every conversation as unpriced translation will feel rigid to work with. The target is unpriced re-explanation forced by a broken record, not conversation as such. Those are different things and it is worth being able to tell them apart.

And the fix moves work forward into the sale, where it is least welcome. Attaching assumptions, thresholds and named approvers to a proposal takes longer than writing a proposal, and it asks a sales team to accept structural constraints at the exact moment they are trying to remove friction from a deal. That resistance is rational and it is the main reason this change fails. If the commercial leadership will not hold that line, the rest does not follow.

The position, restated: leakage is the price of re-narration, it is booked far from where it is caused, and it is fixed at the two hand-overs rather than at the invoice. Start by tracing one recent write-off back to the boundary where the information was lost. It is usually a shorter trip than expected.