Scope creep is not a delivery problem

Scope creep is treated as a delivery failure. It is a commercial control failure that happens to become visible during delivery.

The difference is not semantic. If scope creep is a delivery problem, the remedy is a stricter project manager. If it is a commercial control problem, a stricter project manager will hold the line for a while, lose it during the first difficult client conversation, and the firm will conclude that the person was the problem.

The mistaken diagnosis

In most professional services teams the sequence is the same. A change is discussed in a meeting. Someone says it should be fine. It gets captured in a note, or in nobody’s notes, and it gets delivered anyway. Finance inherits the consequence a month later.

Nothing in that sequence is a delivery error. The work was done, the client was served, the relationship was protected. What did not happen is the commercial half of the decision: the change was never priced, never approved against the contract, and never attached to anything that would carry it into billing.

So the firm agreed to do more work and, separately and silently, agreed to do it for free. Only one of those two decisions was made consciously.

What it looks like downstream

The symptoms are recognisable and they are usually filed as unrelated.

Unbilled work described as goodwill, which is a generous name for a decision nobody made. Margin drift that nobody can account for, because the events that caused it were not recorded as events. Month-end debates about what is chargeable, conducted between people who were not in the meeting where it was agreed. Client disputes, because there is no clear chain of agreement to point at when the invoice arrives.

The last one is the most damaging and the least understood. A client disputing a variation is not usually disputing that the work was valuable. They are disputing that they agreed to pay for it, and in the absence of a record, they are frequently right to.

The mechanism

Controlling scope means making the commercial half of a change as routine as the delivery half. Four steps, in order.

Log the change on the engagement — scope, effort, timeline and the margin effect — at the moment it is raised, not at the point somebody notices the overrun.

Price it against contract rules and rate cards, so the number comes from the agreement rather than from a judgement call made under relationship pressure.

Route it for approval and client sign-off, internally and externally, as a recorded state rather than an email exchange that has to be found later.

Lock it into delivery and billing, so approved work becomes billable and unapproved work cannot quietly enter WIP.

That last step is the one that does the work. If unapproved scope cannot reach WIP, the decision to absorb it has to be taken by a person, on the record, at the time — which is all that governing scope actually means. Whether that discipline held is legible afterwards in the WIP position, which is the difference between a WIP register and a WIP argument.

The rules a variation is priced against are the ones set at the quote, covered on our scoping and quoting page.

What this costs, and where it does not apply

This is friction, deliberately introduced. Somebody has to raise a variation before doing work that sits outside the agreed scope, which is slower than saying yes and is occasionally worse for the relationship in the moment. A firm whose culture is to say yes first and reconcile afterwards will feel that as a loss, and it is a real one — the compensation is that the reconciliation stops happening at the client’s expense a month later.

It also has a floor. On small engagements with a single approver and a stable scope, formal change control costs more than the leakage it prevents, and imposing it produces ceremony rather than control. This matters where engagements are long enough for scope to move, and numerous enough that no one person holds the whole picture.

And it depends on something the process cannot supply. If leadership overrides the gate whenever a significant client pushes back, the gate teaches the team that it is decorative, and the second override costs more credibility than the first saved in goodwill.

Free is a decision, not a default

None of this argues that firms should never absorb work. Absorbing a small change to protect a valuable relationship is often the right commercial call, and a firm that never does it will be an unpleasant firm to buy from.

The argument is narrower. That call should be made by someone with the authority to make it, with the price in front of them, and recorded so the pattern is visible when the same client asks again. A firm that has knowingly given away four variations on an account is in a strong position at renewal. A firm that has unknowingly given away four variations has a margin problem it will attribute to delivery.

Where to start

Take the three engagements that finished furthest from their expected margin last quarter and reconstruct the changes each one absorbed. Not to allocate blame — the point is to find out how many of them exist in writing anywhere, and who, if anyone, priced them.

If the work has changed, the commercials have to change with it. That is not a delivery instruction. It is a commercial rule that delivery needs the system to enforce, so holding the line does not depend on one person’s willingness to have an uncomfortable conversation. Our delivery and time capture page covers where approved change lands once it is priced.