What a change order really costs

What the document actually is

A change order is an amendment to a construction contract. It states a change in scope, a change in price, and, when it is written properly, a change in the completion date.

That third element is the one most often missing, and its absence is the source of most change order disputes. A change agreed on price alone leaves the schedule consequence unallocated, which means it is argued about later, when both parties have an interest in a different answer.

The four costs, of which the price is one

The direct cost is the work itself: materials and labour for the thing being changed. It is the number on the document and it is usually the smallest of the four.

The removal cost is what has to be undone. A change made before an activity starts has none. A change made after has the cost of reversing completed work, and that work was paid for once already.

The sequence cost is the disruption to everything downstream. A change that holds up one trade holds up the trades that follow it, and the size of this cost depends entirely on where in the dependency chain the change lands and on whether the affected trades can be rescheduled without losing their place.

The decision cost is the least visible and often the largest. While a change is being priced and considered, work in that area stops, because nobody proceeds on a detail that is under discussion. A change that takes two weeks to decide costs two weeks of stopped work in that zone whether or not it is ultimately made.

Why the same change gets more expensive every month

Take one example and move it through the programme: relocating a bathroom window.

During design, it is a drawing revision. During framing, it is a framing change and a revised window order, and if the window has already been ordered it is a new window. After sheathing and the water-resistive barrier, it is a framing change, a new opening, a repair to the drainage plane and its flashings, and a patch to the cladding if that has started. After drywall it is all of the above plus interior repair, and after tiling it is all of that plus a tiled wall.

Nothing about the desired outcome changed. The cost rose because more work had to be undone and more trades had to return. That is the general shape of every change on every project, and it is why the useful question about a change is not what does it cost but what has already been built that depends on it.

The change orders that are not really changes

Two categories are worth separating, because they are frequently conflated in the same conversation and they have different causes.

A scope change is a decision to have something different. The owner wants a different thing than the contract described, and paying for it is uncontroversial in principle.

A discovery is different: something the contract assumed turns out not to be true. Rock where soil was expected, a utility in an unexpected position, a condition inside an existing structure that nobody could see. That is not a change in what the owner wants, it is a change in what the job is, and how it is priced depends on who carried the risk of that unknown under the contract. This is where contract type stops being an abstraction, and it is the most common source of genuine dispute rather than mere disappointment.

The third category, and the one worth naming plainly, is the change that exists because the documents were incomplete. If the drawings did not say what colour the cabinets were, choosing a colour is not a change, it is a selection the contract always required. A fixed price signed against an incomplete design generates a stream of these, each individually justifiable, and the sum of them is the difference between the contract price and what the project actually cost.

Deciding about one, in the moment

Four questions, in this order, and they take a few minutes.

Is this a scope change, a discovery, or a gap in the documents. That determines who is responsible before it determines who pays.

What has already been built that this affects. That is the removal cost, and it is usually visible on site.

Which trades does this hold up, and can they be rescheduled without losing their slot. That is the sequence cost, and only the builder can answer it.

And can this wait until after completion. Some changes genuinely can, and a change made after the building is finished is a separate small project rather than a disruption to a large one. That option is almost never presented and is sometimes the right answer.

The change that goes the other way

Change orders are assumed to add money. Some remove it, and those deserve more scrutiny than they get, because a credit is priced by the party giving it up.

Deleting scope late in a job rarely returns what it appeared to cost. The materials may already be ordered or delivered. The subcontractor has reserved the time and priced the job as a whole rather than as a set of separable pieces. And the overhead the deleted work was carrying does not disappear, it redistributes onto everything else.

So a credit for removing work is typically a fraction of what that work contributed to the contract price, and that is not necessarily sharp practice. It is what the arithmetic actually produces. The place to be alert is a credit offered without a breakdown, since the breakdown is the only way to see whether the deduction reflects avoided cost or a round number chosen to end the conversation.

The same logic gives a useful rule for substitutions. Swapping a specified item for a cheaper one mid-job saves the difference in the item and almost nothing else, while introducing a compatibility risk somebody has to check. Substitutions that save real money are made at specification stage, which is before the contract, which is the same conclusion every other section of this piece reaches.

The number worth watching

On a project run properly, the cumulative value of change orders is a measure of how complete the documents were when the contract was signed. A job with very few changes was well specified. A job with a long list was not, whatever anybody says about the reasons.

That is worth tracking as it goes rather than reading afterwards, because the pattern reveals itself early. Three change orders in the first month, all resolving things the drawings did not say, is a reliable signal that the rest of the job will produce many more, and the useful response is to close the remaining open decisions rather than to argue about the ones already priced.

Owner-built projects are not exempt from this and are frequently worse, since the owner acting as general contractor is also the party who decided the drawings were complete enough. [1] The discipline that prevents it is the same in both cases, and it is described on The Build Sheet under planning: close decisions before they become expensive, and know which door each one closes.