A practical guide to delivery operations

Delivery operations is the function that keeps client work honest: knowing what was promised, what is happening, who is doing it, what it is costing and what to do when those stop agreeing with each other. In most agencies and consultancies it exists, but it is spread across a delivery lead, a resourcing spreadsheet, a finance export and a weekly meeting.

The test: four questions, answered today

A delivery operations function is working if it can answer these on any given afternoon, without someone spending a morning assembling the answer.

1. Which projects are not going to land where the plan says, and what is driving that?
2. Where will we be short of the right people in the next few weeks?
3. Which projects are losing margin, and when did that start?
4. What extra work have we taken on that nobody has priced?

If any of those takes more than a few minutes, the problem is usually not effort or discipline. It is that the information needed to answer them sits in separate systems that never compare notes.

Why the reporting cycle hides problems

Traditional delivery reporting is retrospective by design. Time is approved after the week. Costs are reconciled after the month. Status is described in a meeting that happens on a cadence rather than when something changes. Every one of those steps is reasonable, and together they mean bad news arrives once the options have narrowed.

The alternative is not more reporting. It is shortening the distance between a change happening and someone understanding what it implies.

The five things worth connecting first

• Scope: what was agreed, kept where delivery can see it rather than in a signed document nobody opens again.
• Plan: phases, tasks, owners and estimated effort, maintained as the working plan rather than a proposal artefact.
• People: demand derived from the plan, measured against real working patterns rather than an assumed week.
• Time: actual effort captured close to when it happens, so it can inform delivery rather than only billing.
• Commercials: budget, rates, cost and forecast margin reading from the same records as the plan and the time.

Operating habits that make it stick

Tooling helps, but a few habits do most of the work. Keep estimates on tasks even when they are rough, because demand cannot be derived from nothing. Log time weekly rather than monthly, because the cost position is only as current as the effort behind it. Route extra work through a decision instead of absorbing it, so the commercial baseline stays true. And review forecasts, not just status: where a project is heading is more actionable than where it has been.

What good looks like

In a delivery operation that is working, the surprises get smaller. Problems arrive as "this is drifting and here are the options" rather than "this overran". Resourcing conversations happen a few weeks out instead of on the Monday. And scope conversations with clients are commercial discussions rather than apologies.