Running late and overruns — what to say, and how to stop the repeat

One rule covers almost every delay that is your own fault: tell the client early, and give a new time in the same message. The hard part is rarely the wording — it is remembering that a delay you already know about is one you can still get ahead of, instead of one the client discovers by waiting on an empty driveway.

The four delays

Nearly every delay you cause, or absorb, sorts into one of four kinds, and knowing which one you’re looking at decides what you say next:

• Running late today. Traffic, the last job overran a little, or you simply misjudged the morning — you are still coming, just later than promised.
• An overrun into another day. The job itself is bigger than expected, and it will not be finished by the time you said it would.
• A postponement before the day. You already know, ahead of time, that the booked day will not work, and you are the one moving it.
• A block you do not control. Weather that stops outdoor work, a part that has not arrived from a supplier, or another trade on the same job who has not finished theirs.

None of these is the same as a client asking to move a booked job — that one has its own policy, and cancellations and no-shows covers the notice window and the fee. This page is about the day where you are the one who has to change the plan.

What to offer, and who pays for it

Whichever of the four you are facing, the message that actually works has the same two parts: what happened, in one plain sentence, and a real new time, in the same message. A vague explanation with no new time attached leaves the client to chase you for the part that matters.

• Say it before the client has to ask. A delay you flag yourself reads as a business staying on top of things. The same delay, discovered when the client calls to ask where you are, reads as being let down.
• Name a specific new time, not "soon." A day and a rough time is a plan the client can work around. "I'll be in touch" is not.
• You absorb the cost of the wasted slot. When the delay is yours, the slot you cost the client is yours to make good, not theirs to eat — the same logic runs the opposite way from a client's late cancellation, where you keep the fee you were promised.
• A small goodwill gesture, proportionate to the inconvenience. A discount, a free extra, or waiving the call-out fee on the return visit, offered before the client asks, reads as taking ownership. Save it for a delay that genuinely cost the client something, not for ten minutes on a driveway.

The wording is where most people stall, usually because a delay feels like it needs an explanation long enough to justify itself. It doesn’t. The delay and overrun scripts cover all four kinds — running late today, an overrun into tomorrow, a postponement, and each of the blocks you don’t control — plus the apology and the goodwill offer once a slot has already been missed.

When nobody is at fault

Weather, a late part from a supplier, or another trade who has not cleared the way — none of it is a mistake you made, but the client is still waiting on a date. Naming the actual cause plainly, rather than a vague "something came up," is worth doing, but it is not a substitute for giving a new time. That part of the message stays yours to offer no matter who or what caused the hold-up.

Your terms should already say, in plain words, that a genuine force majeure delay — weather, a supplier, another trade — does not carry the same fee wording as your own late cancellation policy. Stating it up front, before it is ever needed, is what keeps a real delay from turning into an argument about blame.

Stopping the repeat

A delay that keeps recurring is usually a scheduling habit, not bad luck. Four habits catch most of it before it starts:

• Hold a buffer between jobs. A day booked hour to hour has nowhere for an ordinary overrun to go; a day with a gap in it absorbs one without knocking on to the next client.
• Confirm the day before. A short day-before message is also the moment you notice a job that is already at risk of running long, while there is still time to warn the client.
• Check the lead time before you book. A part, a material, or another trade's own schedule has a lead time attached — book against it, not against hope that it arrives in time.
• Never book the slot you cannot afford to lose. If a delay on this job would cascade into every booking after it, the day was already too tight before anything went wrong.

A rushed morning can look fine in your head and still not fit on paper once travel and time on site are actually added up. The free day plan check runs that arithmetic for you before you say yes to one more job, the same way it helps with a diary clash between two of your own bookings.

Common questions

Do I always owe an apology, or a discount, for running late?

An apology, yes — always say sorry for the wait. A discount, not always. Save a goodwill gesture for a delay that genuinely cost the client something, such as a missed morning or a knock-on delay to their own plans, not for ten minutes on a driveway.

What do I say when the delay genuinely is not my fault?

Name the actual cause plainly — weather, a supplier, another trade — rather than a vague "something came up." Naming the cause explains why it happened; it does not replace giving the client a new date. That part is still yours to offer, whoever caused the delay.

How is this different from a client cancelling on me?

A client cancelling or moving a booking is their decision, and cancellations and no-shows covers the policy for it. This page is about the reverse: you are the one who has to move, delay, or cannot finish a job, and the client did nothing to cause it.

Should I charge my own cancellation fee back to myself?

No — a cancellation fee compensates you for a client’s late notice. When the delay is yours, the honest response runs the other way: absorb the cost of the wasted slot, and consider a small goodwill gesture if the delay genuinely cost the client something.

How much slack should a day actually carry?

Enough that an ordinary overrun does not cascade into every job after it. Look back at how often a job actually overruns, and by how much, then build a buffer close to that figure — a trade with more variable jobs needs more slack than one with short, predictable visits.