What is Change Control?

Every engagement has a moment when the work stops matching the contract. A request in a status call, a dependency that moved, an assumption that turned out to be optimistic. Change control is the discipline that decides what happens next: whether the drift becomes a recorded, priced decision, or disappears into the delivery team’s week and resurfaces months later as a margin nobody can explain.

What change control actually is

Change control is the agreed process for handling variation deliberately. The essential loop has four steps, and every functioning version of it, however formal or informal, contains all four.

• Raise. The change is surfaced and written down, so it exists somewhere other than a conversation. This is the step most often skipped, because raising a change feels like friction against a relationship that is going well.
• Price. Someone estimates what the change costs: effort, elapsed time, dependencies, risk. Pricing does not commit anyone to charging for it. It commits the firm to knowing what it is about to spend.
• Approve. A person with authority over the commercial outcome decides: charge for it, trade it against something already in scope, absorb it knowingly, or decline it. The client approves their side too, which is what makes the variation binding rather than assumed.
• Record. The decision is written into the engagement’s record, so the current scope, price and timeline remain a fact rather than a memory.

Notice what the loop does not do. It does not prevent change, slow the client down or forbid generosity. Absorbing a request for free is a perfectly legitimate outcome of change control. The difference is that it happens on purpose, at a known price, decided by someone entitled to spend that money.

Change control and change orders are not the same thing

The two terms travel together and are worth separating. Change control is the process; the change order is the instrument the process produces: the signed document recording one approved variation, its price and its impact on the schedule. The distinction matters because firms routinely mistake owning the instrument for running the process. A beautiful change-order template achieves nothing if requests are absorbed in the delivery flow before anyone raises them. The template is the last step of a discipline, not a substitute for it. A firm with strong change control and rough paperwork is in far better shape than a firm with polished paperwork and no habit of raising change while it is still cheap to decide.

A margin problem before it is a legal one

Weak change control is usually filed as a contractual risk: undocumented variations, no signed basis for extra billing, exposure in a dispute. All true, and all occasional. The continuous damage is commercial. Weak change control is a delivery-margin problem long before it is a legal one, because unmanaged change does not stop the work, it only stops the record. When no process catches a request, the work still happens. The cost is incurred, the revenue is not, and the engagement’s economics move without anyone deciding they should. This is scope creep seen from the supplier’s side, and repeated across an engagement it becomes one of the main sources of the gap between sold and delivered margin. Most engagements never end up in a legal dispute. Nearly all of them end up at closeout, where the absorbed changes are finally counted, too late to do anything about them.

One concrete example

Clearly illustrative, with no customer implied. A firm of a few hundred people runs a year-long implementation. In month two the client asks for an extra integration in a stand-up; the team, ahead of schedule, quietly builds it. In month five a regulatory tweak on the client side changes a reporting format; two weeks of rework, absorbed, since nobody wants to invoice a client for their regulator. In month eight the client’s new operations lead asks for a parallel-run period that was never in scope. Each request was reasonable, each absorption was kind, and none of the three passed through anything resembling raise, price, approve, record. At closeout the engagement lands well under its sold margin, the account team calls it a tough delivery, and the honest answer is that the firm made three unpriced commercial decisions and let the people least placed to price them make all three.

The decision-intelligence reading

Look again at the four steps and change control reveals what it really is: a decision process wearing project-management clothes. Raise is evidence capture. Price is options priced against their cost. Approve is a human choice made by someone accountable. Record is the decision written down with who made it and why. That is the anatomy of any governed decision, applied to the small commercial choices that engagements generate weekly, and it is why decision intelligence treats change control as a native habit rather than an imported bureaucracy. The firms that keep scope honest are not the ones with the strictest contracts. They are the ones where a request arriving in a status call reliably becomes a recorded decision while there is still time to choose, and where the running ledger of those decisions means the closeout review confirms what everyone already knew, instead of revealing it.