The Lapsed Maintenance Customer SOP

Purpose

To define how this shop finds, works, and dispositions customers who are on a maintenance rhythm and have missed their due date. A missed maintenance visit is the earliest and cheapest churn signal a service business gets, and it is the one most shops never look at, because nothing happens when a visit does not occur. No phone rings, no complaint arrives, and the schedule quietly gets easier. Twelve to eighteen months later the same customer appears on a win-back list, where recovering them costs many times what a phone call would have cost in the first thirty days.

This document covers the interval between the missed due date and the point where a customer stops being lapsed and becomes dormant.

Scope

Applies to every customer with an expected recurring service, whether that expectation comes from a signed agreement, an enrolled plan, or simply an established pattern of visits at a regular interval.

Does not apply to customers with no recurring pattern, active jobs in progress, or customers already sitting on the dormant or lost list. Those are governed by the reconnection and win-back references at the end of this document.

Definitions. Due date is the date the next service should occur, derived from the last completed visit plus the service interval for that customer's service type. Lapsed means past the due date with no visit and no scheduled appointment. Dormant means past 180 days lapsed, at which point the customer leaves this process.

Roles and responsibilities

Procedure

1. Set the due date rule per service type before anything else

For each service you deliver on a rhythm, write down the interval in months and the tolerance window. A customer is not lapsed the day after the anniversary if your tolerance is four weeks. Publish the intervals so office and field agree, because a technician telling a customer "you are fine until the spring" while the office is calling them lapsed destroys the credibility of the whole process.

Where the interval genuinely depends on conditions, for example heavy use, harsh environment, or an asset past its typical service life, record the customer-specific interval as a field on the record rather than leaving it to memory. Otherwise the list will call the same heavy-use customer lapsed on the standard interval every cycle and staff will learn to ignore the list.

2. Generate the lapsed list every week, at a fixed time

This runs off service history and a calendar and needs no software beyond whatever holds your job records. Weekly, pull every customer whose due date has passed with no completed visit and no booked appointment, and stamp each with days past due.

Weekly, not monthly. A monthly pull means the average bucket A customer is already three weeks into their lapse before anyone contacts them, which is most of the window in which recovery is nearly free. It also produces a large batch that gets deferred, where a weekly pull produces a handful that gets worked.

3. Triage into four aging buckets and work each with its own action

Bucket A is where the money is and it is the bucket that gets skipped. A customer 12 days past due has not made any decision about you. They missed a message, they were travelling, the reminder went to an old number. The recovery is a two-minute call. By bucket C the customer has lived through most of a cycle without you and has usually formed an opinion, which is a much harder conversation with a much lower success rate.

No win-back offer is ever made to a bucket A or B customer. They have not left. Offering a concession to someone who was merely late teaches your most reliable customers that missing a date is rewarded, and it is the fastest way to convert a healthy maintenance base into a discount-negotiating one.

4. Use the two-named-dates rule on every reschedule contact

Never ask "when would work for you." Offer two specific dates with time windows and let them pick or counter. An open question puts the scheduling work on the customer and produces "let me get back to you," which is the most common way a bucket A customer becomes a bucket B customer.

This is why the calendar owner holds a protected reschedule slot each week. Two named dates you cannot actually honour is worse than an open question.

5. Record a disposition code on every contact, without exception

One code per contact, from a fixed list. Suggested set: RESCHEDULED, DECLINED THIS CYCLE, DECLINED ENDED, UNREACHABLE, MOVED OR SOLD, LOST TO COMPETITOR, LOST SERVICE ISSUE, ASSET REMOVED.

The codes are the only output of this process that has value beyond the individual customer. A quarter of DECLINED THIS CYCLE codes clustered in one month tells you your due dates fall in a season your customers cannot accommodate. A run of UNREACHABLE codes tells you the contact data is stale and the problem is in your onboarding capture, not your calling. Without codes you have a call log and no information.

6. Handle plan and agreement members differently

A customer with a prepaid or contracted maintenance visit is not simply lapsed; they are owed something you have taken money for. That changes three things. Their contact happens at bucket A without waiting for the weekly cycle. The unperformed visit is tracked as an outstanding obligation rather than a scheduling gap. And the conversation is not a request but a delivery: you are trying to give them something they already bought.

Never allow a contracted visit to age past the term without a documented refusal in writing. An unperformed prepaid visit at renewal time is simultaneously the customer's strongest argument for cancelling and a legitimate grievance, and it is the most common thing that turns a plan renewal into a refund conversation.

7. Hand off cleanly at 180 days

At bucket D the customer leaves this process. Record the reason code, stop the maintenance reminder cadence so they do not keep receiving due notices for a service they are no longer receiving, and move the record onto the dormant list where it will be worked once a year on a different footing.

The stopping of the cadence is the part shops forget. A customer who declined and then keeps getting due notices for two more years learns that your messages are automated and stops reading all of them, including the safety notice you may one day need them to read.

8. Review three numbers weekly

At a standing weekly slot, the office lead and the owner look at exactly three figures: how many customers entered bucket A this week, how many moved from A to B without being contacted, and the disposition mix for the week.

The second figure is the one that matters. A customer aging from A to B uncontacted is a process failure, not a customer decision, and it is entirely inside your control. If that number is not near zero, nothing else in this document will work.

Worked example

All figures illustrative. A shop has about 420 customers on an annual maintenance rhythm, so roughly 35 due each month.

The lapse volume. If about 78% of each month's due customers get their visit inside the tolerance window, roughly 22% lapse, which is about 8 customers a month, or on the order of 92 across a year.

The recovery. With this SOP running, suppose about 70% of those 92 are recovered before they reach 90 days past due. That is about 64 recovered and about 28 aging past bucket B toward dormant. The shop's historical recovery before adopting the process was around 35%, or about 32 of the 92. So the process retains roughly 32 more maintenance customers a year than the shop was retaining before.

The cost. Those 92 lapses take an average of about 2.2 contacts each, so around 202 contacts a year, at roughly 0.15 hours per contact including the record work. That is about 30 hours a year, or a bit under 35 minutes a week of office time.

The return, read honestly. 32 additional retained customers at about 1.2 billable hours a year each is about 38 billable hours a year, against roughly 30 hours of office time. In pure first-year hours that is a ratio of about 1.25 to 1, which on its own is unimpressive, and a shop that judges the process on that number alone will drop it.

The reason to run it anyway is what the ratio leaves out. The 30 hours is an annual cost against the entire lapsed population; the 32 retained customers stay on the rhythm for years, so the second year's 38 hours arrives with no new investment in those customers. And maintenance customers are the population that generates your repair and replacement work, which does not appear anywhere in the 38 hours. Losing 32 maintenance customers a year is losing the front end of your repair pipeline, and that consequence surfaces two to four years later, at which point it looks like a demand problem rather than a process failure.

Watch out for

A lapse list that includes customers who were never on a rhythm. Padding the list with episodic customers makes it too long to work, and staff respond by working the top of the list rather than the aging order. Audit the list membership quarterly.

Technicians extending intervals verbally in the field. A technician telling a customer "you probably only need this every other year" changes that customer's due date and nobody records it. Either the interval field gets updated on the spot or the technician does not have that authority. Both are workable policies; leaving it undefined is not.

The process becoming a performance review. The A-to-B uncontacted figure exists to find broken workflow: a reminder going to a dead number, a due-date rule that is wrong, a week where the office was covering a sick day. If it becomes a stick used on one person, the reliable adaptation is to record contacts that did not happen, and the number stops meaning anything.

A closed gap reopening silently. A customer recovered from bucket B who then misses the following cycle is a different situation from a first-time lapse and should not be worked as one. Flag repeat lapses at the point of triage; two consecutive lapses is a relationship question for the owner, not a scheduling question for the office.

Nobody turning on the disposition log. The most common outcome of adopting this document is that the calling starts and the coding does not, because the coding has no immediate payoff. Six months later the shop has recovered customers and still cannot say why anyone lapsed. Make the code a required field on the contact record, not a note.

References

• U.S. Small Business Administration (SBA), customer retention practices for small service businesses
• Manufacturer maintenance-interval documentation and trade-standard service intervals for the equipment you service
• See related: The Renewal Cadence That Reduces Churn; The Win-Back Offer Design SOP; The Communication Cadence Between Jobs; The Customer Health Score a Small Shop Can Actually Keep