The Change That Should Have Been a Change Order

Why this matters

Shops are good at papering the change that looks like a change: extra work, a new part, a bigger number. The ones that go unpapered are the ones with nothing new in them. A substitution at the same price. A resequence to suit the customer's tenant. A two-week hold while somebody decides. A deletion. None of those feel like a sale, so nobody writes them up, and then the shop absorbs the hours and, more expensively, silently accepts responsibility for a completion date it no longer controls. The money in these changes sits in time and in risk allocation, not in parts, which is exactly why the usual test misses them.

The test that misses them

The instinctive test is novelty: is there work here that was not in the quote? That question is easy to answer and it is the wrong question. A change order is not a receipt for new work. It is the instrument that re-states who carries a specific consequence after something moved.

Use an impact test instead, and measure against the quoted baseline rather than against what you feel like the job should be:

A change gets papered when, compared to the quoted baseline, it moves labor by more than 0.5 hour, OR moves the completion date by more than one working day, OR changes the material set, OR shifts who carries a named risk. Any one leg is enough. The unit of analysis is one occurrence, not one job, and approved labor is written in 0.25-hour steps.

Novelty is not on that list. Work can be brand new and fail every leg. Work can be identical to the quote and trip three.

The five changes with no new work in them

• The equal-price substitution. The customer prefers a different finish, model class, or material that costs the same. No new labor in the install, and possibly a long lead time, a second mobilization, and a completion date you now own without having agreed to it.
• The resequence. Same scope, different order, because a tenant needs a room back or another trade is behind. Resequencing usually costs setup time and sometimes costs a return trip, and it always costs the efficiency you priced into a continuous run.
• The hold. The customer needs a week to decide. Nothing is added or removed. Your crew is committed elsewhere by the time they decide, so the job restarts cold, and every date in your quote is now stale.
• The deletion. Work comes out. Feels like a gift to the customer and a loss to you, and it is neither until somebody works out what the deleted portion was carrying. A deleted item often carries a share of mobilization, setup, and a minimum charge that does not shrink with it.
• The access change. Same work, new conditions: a locked room, a shortened work window, an escort requirement, a change in staging location. This is the one that most reliably breaks an hours estimate while leaving the scope sentence untouched.

Deletions need the arithmetic stated, not just the credit

A deletion is the change most often handled by feel, and feel is wrong in a specific direction. If a quoted job carries one mobilization, one setup, and one round of testing, deleting 20% of the installed items does not remove 20% of the hours. It removes the labor that was specific to those items and none of the shared labor. Credit the item labor and the item material, state in the same line that mobilization, setup, and testing are unchanged because they were priced once for the visit, and you have a defensible credit rather than an argument about whether you shortchanged them.

The corresponding failure is a credit given at the full line rate, which quietly converts a profitable job into a break-even one while looking generous.

Price the remobilization once, in the quote, so the time changes have a number

The reason resequences, holds, and substitutions go unpapered is not only that they do not look like changes. It is that nobody has a figure for them, and inventing one mid-job in front of a customer feels like a penalty. Fix that in the quote by publishing a standing figure for an additional mobilization: the labor to demobilize, return, re-stage, and re-establish the work area, stated in hours and applied per occurrence beyond the number of mobilizations the quote includes.

Two lines do it. State how many mobilizations the price includes, and state what each additional one costs in labor hours. Then a customer-requested hold or resequence has an agreed number attached before anyone needs it, and the change order becomes arithmetic rather than an argument. Shops that do this find the number itself rarely gets challenged; what gets challenged is a figure that first appears after the inconvenience.

Size it from your own data rather than from feel. Pull a handful of jobs that required a second trip and look at what the second trip actually consumed from wheels-rolling to tools-in-hand, and use the median rather than the average, because one bad traffic day distorts a small sample.

One gate, two changes, opposite answers

A crew is quoted 22.0 labor hours over 3 working days on a tenant space. The material list is fixed, the completion date in the quote is the end of working day 3.

Change one: the customer wants a different but equal-price material. Requested at midday on day 1, after 2.5 of the 22.0 hours are done. There is no new work in it. The install labor for the substituted material is identical. The material cost is the same. Run the gate:

• Labor leg: the crew has to demobilize and return, adding 1.0 hour of remobilization. Remaining work is 19.5 hours (22.0 minus the 2.5 already done), so total labor becomes 2.5 plus 19.5 plus 1.0, which is 23.0 hours. That is 1.0 hour over the 22.0 baseline, or about 4.5%. The leg trips at more than 0.5 hour, so it trips.
• Schedule leg: the substituted material lands at the end of working day 7. The crew returns on day 8 and finishes on day 10. Completion moves from day 3 to day 10, a slip of 7 working days. The leg trips.
• Material set: changed. Trips.
• Risk allocation: the delivery date of the substituted material is now a risk sitting on the job, and nobody has said whose it is. Trips.

Four legs out of four, on a change with no new work and no price difference in the material. What gets papered is a change order that adds 1.0 hour, re-states completion as end of day 10, and says in one sentence that the revised date depends on the substituted material arriving by end of day 7 and moves day for day if it does not. That last sentence is the whole reason to write it. Without it, the shop has taken a customer-chosen delay onto its own schedule commitment.

Change two: on day 8 the quoted routing will not clear an obstruction. The tech re-routes using fittings already on the quoted material list. This is genuinely new work, done differently than planned, decided in the field. Run the same gate:

• Labor: 0.3 hour extra. Below the 0.5-hour step, so it does not trip.
• Schedule: no move. Does not trip.
• Material set: unchanged, the fittings were on the list. Does not trip.
• Risk: unchanged. The result meets the same specification.

Zero legs. This is not a change order. It is means and methods, which is your side of the line, and the 0.3 hour is yours to absorb. It still gets a job note and a photograph of the as-built routing, because the next tech and possibly an inspector will need to know the deviation was deliberate and where the line actually runs.

Now look at what a shop running the novelty test does with these two. It papers change two, because there was new work, and the customer reads a change order for 0.3 of an hour as nickel-and-diming. It does not paper change one, because nothing new happened, so the shop eats 1.0 hour and, far worse, is the party standing there on day 10 explaining a slip it never agreed to carry. The cheap loss is the hour. The expensive loss is owning a date that moved for someone else's reason.

The zero-cost change order

The instrument most shops are missing is a change order that changes no money at all. It exists to move a date, restate an assumption, or reassign a risk. It reads: what changed, who asked for it, what it does to the completion date, what it does to the price (nothing), and what the revised date depends on.

Customers accept these easily, because there is no number to argue with. That is the point. The moment there is a number, the conversation becomes a negotiation about the number, and the date buried underneath it goes unread.

Where a contract carries a delay clause, a liquidated damages provision, a notice period for claiming an extension of time, or a no-damages-for-delay term, the consequences of an unpapered slip are set by that contract and by the law of the state whose law the contract says governs, and several of those terms are limited or unenforceable in some states. Read the clause, meet its notice deadline, and have your own attorney tell you what your contract's version actually does. Do not resolve that from an article.

How to check whether this is happening to you

• Pull the jobs that finished late and ask what moved the date. If the reason is a customer decision, a substitution, or a hold, and there is no change order carrying that date, you have found the pattern. Count how many. Compare that count to the number of change orders you issued for added work in the same period, because the ratio is the finding.
• Read your change orders for date language. If they all adjust price and none of them adjust the completion date, you are papering money and absorbing time.
• Look for the deletion with a full-rate credit. Then check whether mobilization and testing hours came out with it. They should not have.
• Ask the crew what they redid this month because someone changed their mind. Techs remember resequences and holds precisely, and they almost never report them, because nothing was added.

References

• Contract terms governing delay, extensions of time, notice periods, and liquidated damages vary by state law and by whether the customer is a consumer or a business; have your own attorney review the change order and delay language you use
• See related: How to Stop a Job From Growing Without a Fight, The Change Order as a Contract Amendment Not a Formality, Why Verbal Approval Is Worth Less Than It Feels