What is Scope Creep?

Almost no services firm loses its margin in one bad quarter. It loses it a favour at a time. An extra report added to the monthly pack because the client asked nicely. An integration nobody scoped because it sounded trivial in the workshop. A weekly call that quietly became a daily one. Scope creep is the name for this slow widening of what a team delivers beyond what was contracted and priced, and it is dangerous precisely because no single instance ever feels worth a difficult conversation.

The standard framing treats scope creep as a delivery problem: weak project management, soft boundaries, a team that cannot say no. The framing is comfortable, and wrong. Boundaries do not erode because delivery people are weak. They erode because the commercial choice hidden inside every request is never surfaced to anyone with the authority to make it.

Scope creep is a decision problem wearing a delivery costume: every absorbed request is a pricing decision, and absorbing it silently means the decision was made by default, by whoever happened to be in the room, at a price of zero.

Every absorbed request is a pricing decision

When a client asks for something outside the agreed scope, the firm faces a genuine commercial choice: charge for it, trade it against something else, invest it in the relationship, or decline it. That is a pricing decision, and pricing decisions have owners. But the request rarely arrives at the owner's desk. It arrives in a stand-up, an email thread or a status call, addressed to a delivery lead whose incentives all point one way: keep the client happy, keep the project moving, avoid friction. So the yes is given where the request landed, and the pricing decision is made by default, without the person, the analysis or the record that a pricing decision deserves.

Two properties make the default expensive. First, absorbed work compounds. What was granted once becomes the baseline, and the next request starts from the new baseline. Scope ratchets; it never creeps back. Second, the erosion is invisible at the moment it happens. Each favour costs hours, not invoices, so nothing shows up until the margin review months later, by which point nobody can say where the margin went, only that it did. Creep can also only be detected against a scope somebody wrote down, which is why a vague statement of work is where it grows fastest.

Goodwill is not the same as unpriced work

The usual defence of absorption is goodwill: we are investing in the relationship. Sometimes that is true. But genuine goodwill has three properties that silent absorption lacks. It is decided, by someone with the authority to spend the firm's money. It is bounded, with a value someone could state. And it is visible, above all to the client, who is told that something of value is being given.

Silently absorbed work has none of these. Nobody decided it, nobody bounded it, and the client never registered it as generosity because it was never presented as having a price. That is the quiet tragedy of scope creep: the firm bears the full cost of generosity and collects none of the relationship credit for it. A gift the recipient does not know is a gift buys nothing.

The three honest ways to say yes

There are only three honest responses to an out-of-scope request, and all three are a yes:

• Price it. Raise a change order. The work is valued, agreed and billed. This is the default that keeps scope and price connected.
• Trade it. Yes to the new request, in exchange for something: an item descoped, a date moved, a deferral to the next phase. Scope is exchanged rather than expanded, and the client learns that scope has weight.
• Gift it knowingly. Sometimes investing free work in a relationship is exactly the right call. But it is a call: made by someone with pricing authority, with a stated value, recorded, and communicated to the client as an investment.

The fourth response, absorbing the request silently, is not a kinder version of the gift. It is the same cost with the decision, the record and the credit removed.

One concrete example

Clearly illustrative, with no customer implied. A firm of a few hundred people runs a managed service priced against an agreed monthly reporting pack. Over a year, small requests accumulate: an extra data cut here, an ad hoc analysis there, a recurring call that was never in scope. Suppose, purely for arithmetic, that the absorbed requests average three a week at half a day of effort each. That is roughly seventy days a year of unpriced delivery on a single account, a meaningful fraction of a full-time person, visible nowhere. Then renewal arrives, and the trap closes. The client's baseline expectation now includes everything that was absorbed, so pricing the renewal at the true cost of service reads to the client as a price increase for the same service. The firm must either surface a year of hidden generosity in one awkward conversation, or bake the loss into another term.

The decision-intelligence angle

Everything above describes decisions that were made without being noticed: hundreds of small pricing calls, each owned by nobody, each recorded nowhere. That is why scope creep belongs to decision intelligence rather than to project hygiene. The remedy is not heroic boundary-setting by delivery teams; it is moving the choice to where it can be seen. A request is surfaced as a decision, its cost is estimated, its owner is named, and the chosen response (priced, traded or gifted) is recorded, so the pattern across accounts becomes visible long before the margin review finds it. Absorbed work is also one of the main tributaries of revenue leakage, and the wider discipline of treating such choices as governed, recorded decisions is set out in Enterprise Decision Intelligence. The firms that handle scope creep well are not the ones that say no most often. They are the ones where every yes was chosen.