How to Handle a Customer Who Only Calls in Emergencies

Why this matters

Every shop has a handful of accounts that never appear on the schedule until something is broken, flooding, dead, or dangerous. The tickets look good on paper because emergency work bills higher, so the pattern goes unexamined for years. What it actually costs shows up somewhere else: in the scheduled jobs you push to absorb the call, in the tech who works a Saturday night, and in the fact that a customer in crisis will call three shops instead of one. Handling this pattern is a lifecycle problem, not a dispatch problem, and it has a procedure.

Step 1: Confirm the pattern from service history, not from memory

Office staff will name the account they remember. Memory over-weights the loudest calls. Pull the ticket history for the account over the last 24 to 36 months and mark each ticket as booked in advance or arrived unscheduled. The working definition of an emergency-only customer: over a trailing 24 months or more, every single ticket arrived unscheduled, and no visit was ever placed on your calendar before the day it happened.

That is a stricter test than "they call us a lot in a panic," and it matters. A customer who books a spring visit and also calls you at midnight in August is a normal maintenance customer having a bad night. Treat them as emergency-only and you will insult a good account. If the history is thin because you only started keeping tickets recently, use what you have and mark the conclusion provisional rather than acting on two data points.

Step 2: Read the cause before you design the fix

Four causes produce the identical pattern, and they need different handling.

• No belief prevention works. They have been sold a maintenance visit before, nothing was found, and the failure happened anyway. They concluded the visit was theater.
• No room in the budget for anything not on fire. Common in rental portfolios and in households running tight. They are not skeptical, they are rationing.
• You are the second shop. Somebody else has the planned work and you get the calls that other shop cannot cover at 9pm. This is the one owners misread most often.
• Nobody lives there. Absentee owner, seasonal property, or a manager covering a site they never visit. Prevention requires someone on site to notice a warning sign, and there is nobody.

You find out by asking on a normal weekday, not during a call-out. "When something like this comes up outside of hours, who else do you have in the rotation?" and "Has anyone ever done a scheduled check on this for you?" Two questions, thirty seconds, and they sort the four causes cleanly.

Step 3: Price the pattern correctly while it is still the pattern

Before you try to convert anyone, make sure the current arrangement is not being subsidized. After-hours and same-day work carries real cost that a normal ticket does not: a tech pulled off rest, a truck rolled outside routed hours, and the scheduled customers who get pushed. If your after-hours rate is not distinctly above your booked rate, you have built a hotline and priced it like an appointment.

This is also the step where discipline gets tested. Discounting the emergency rate for a good long-time customer feels generous and teaches exactly the wrong lesson: it makes the unplanned call the cheapest way to buy your time. Keep the rate, and put the generosity somewhere that reinforces the behavior you want, such as crediting a portion of the emergency call toward a scheduled visit booked within 30 days.

Safety belongs in this step too, because after-hours calls are where it degrades. Set a hard rule that a tech going alone to an unfamiliar property after dark texts dispatch or the on-call lead on arrival and again on departure, with the address. On any call where the customer reports a gas smell, the instruction to the customer on the phone is specific: everyone leaves the building right now, do not touch any switch, do not turn lights on or off, do not use a phone inside, call the gas utility from outside and stay out until they clear it. Your tech does not walk into an unvented gas odor to diagnose it, and neither does the customer.

Step 4: Pick the conversion moment, which is narrow

The window to convert an emergency-only customer opens about 3 days after the emergency is resolved and closes around day 10. Earlier than that and you are selling to someone still in the middle of it, who will agree to anything and mean none of it. Later than about two weeks and the memory of the 2am ceiling has faded, the bill has been paid and mildly resented, and you are back to a cold offer.

Do not use the moment at the end of the call-out itself. A tech packing tools at 1am saying "you should really get on a plan" reads as an upsell attached to the worst night of their month, and it is the single most common way this conversation gets poisoned for a year.

Step 5: Make the ask about the failure they just had

The ask is not "would you like to join our maintenance program." That invites a comparison to every other program they have declined. The ask is a specific claim about the specific failure, and it is honest about the odds.

"I want to look at the last three times we came out for you. Two of the three started as something that shows up on a scheduled check. I am not going to tell you a visit would have caught everything, because the third one would have failed no matter what. But two out of three is worth a conversation. Can I put you down for a check in April and we see what it finds?"

That works because it concedes the limit. A shop that claims prevention catches everything is disbelieved by anyone who has had a preventable-sounding failure happen anyway. A shop that says two of three, and names the one it would have missed, is credible.

If the cause from Step 2 was "you are the second shop," the ask changes entirely. You are not competing on prevention, you are asking for a trial of the planned work: "You clearly have someone for the routine side. If they are ever short, I would take one of those visits this year and you can compare." If the cause was "nobody lives there," the ask is a remote-condition arrangement instead: two visits a year plus a written condition report the owner receives, because the value you are selling is eyes on a property nobody is watching.

Step 6: Close with a date, not an intention

An agreement in principle is worth nothing on this account type. Book the visit before you leave the call or hang up the phone, put it in the calendar in front of them, and send the confirmation the same day. If they will not take a date, they have declined, and you should record that rather than carry a phantom in your head as a converted customer.

Step 7: When they will not convert, set standing terms instead

Some accounts are permanently emergency-only, and that is a legitimate arrangement as long as both sides know the terms. Say them out loud once, and write them into the file so whoever answers the phone next time says the same thing: what your realistic response commitment is outside of hours, that the after-hours rate applies, and that booked customers get the slot first when both call at once. That last one is not a threat, it is a description, and stating it plainly is what stops the resentment that builds when it happens silently.

Worked example: one account over five years

A shop pulled 38 months of history on a small commercial site. 5 tickets, all 5 unscheduled, four of them same-day or after-hours and one on a weekend. That is one ticket every 7.6 months on average, all of it unplanned.

The emergency tickets averaged 2.4 times the shop's median scheduled ticket, which is why nobody had looked at the account critically. Two other numbers had never been counted. First, those 5 call-outs displaced 7 already-booked appointments over the same 38 months, about 2.2 pushed appointments a year. Second, on review, 2 of the 5 failures traced to a condition a routine check would plausibly have flagged, so 40% of the emergencies were arguably preventable and 60% were not.

The shop ran Step 5 at day 6 after the fifth call-out, using the two-of-five framing, and booked two visits a year. Over the following 24 months the account produced 4 scheduled visits and 1 emergency, 5 tickets in 24 months, which is one ticket every 4.8 months. Annualized, ticket count went from about 1.6 a year to 2.5 a year, an increase of roughly 58% in tickets per year on that account.

Annual billing did not follow the ticket count, and this is the part worth sitting with. Measured in median-ticket-units, the emergency-only years ran about 1.6 tickets a year at 2.4 units each, so roughly 3.8 units a year. The converted years ran 2 scheduled visits a year at about 0.6 units each, plus found-and-approved repair work averaging about 1.1 units on half the visits, plus 0.5 emergencies a year at 2.4 units. That totals about 3.5 units a year, which is roughly 8% lower annual billing on that one account than the emergency-only pattern produced.

So the conversion did not grow the account. What it did was move 2.5 tickets a year onto a calendar the shop controls, and cut displaced appointments from about 2.2 a year to about 1.0 a year. The scheduled visits landed in soft weeks. The customer stopped calling two other shops during a crisis, because there was now a shop that already knew the site.

That is the honest case for this work: you are not converting an emergency-only customer for the revenue, you are converting them for the calendar. If you promise your team a revenue jump from this program, the numbers will not back you up and they will stop running it in year two.

What would change the recommendation

If your shop is capacity-constrained and turning work away, converting emergency-only accounts is a low priority and may be negative. You would be adding scheduled visits at a lower unit value while the emergency work you were already getting fills the same hours at a higher one. Run this program when you have soft weeks to fill, not when you are booked out.

If the account is a single-site rental with an owner who genuinely cannot fund prevention, do not run Step 5 at all. Run Step 7, price it correctly, and stop spending office time on a conversion that will not happen.

How to verify you got this right

Six months in, count three things per converted account: tickets that arrived unscheduled, appointments you pushed to absorb an emergency, and whether the scheduled visits actually happened or quietly slipped. The third is the one that fails first. A converted account that has skipped two consecutive scheduled visits has reverted, and it should go back on the list rather than sit in the file marked converted.

References

• Trade-standard practice for after-hours rate structures and on-call rotations
• OSHA general industry guidance on working alone and site check-in procedures for employees
• See related: The One-Time Customer Versus the Relationship Customer, The Annual Tune-Up Visit as a Retention Anchor, Customer Membership Programs