The customer list nobody is using

Sitting in the invoicing system of almost every established service business is a list of everyone who has ever paid them, and in most cases nobody has ever contacted a single one of those people again. It is the cheapest revenue the company has access to, it requires no marketing spend and no new capability, and the reason it goes unused is not that owners disagree. It is that reactivation is nobody's job.

What is actually in there

A fifteen-year-old trade business has served thousands of customers. Some were one-off jobs, plenty were repeat, and a meaningful share had a piece of equipment installed that has a service life and a maintenance interval. Every one of those records has a name, an address, a phone number, and a description of work done, sitting in accounting software that was never intended to be a customer database and is functioning as one anyway.

Four distinct things are buried in that list, and they are worth separating because they need different treatment.

1. Equipment that is due

   Anything installed with an inspection or maintenance cycle attached. The revenue is contractually or legally implied and simply not being collected, because no system knows when each unit falls due. This is the largest and most predictable piece.

2. Customers who lapsed without deciding to

   People who used you twice, then had no need for three years, and have not thought about you since. They did not leave, and the next time they need the service they will search rather than remember, which is a competitive loss disguised as a passage of time.

3. Quotes that were never chased

   Priced work that went out and got no answer. Some of those went to a competitor and some went nowhere at all, because the customer got busy and the job never happened. The second group is still live and nobody has ever found out which is which.

4. The second site nobody ever quoted

   Commercial customers who own more than one property and only ever had you at one of them, usually because that is where the thing broke. Nobody has ever asked about the others, and the relationship needed to do so already exists.

The first is the one to start with, because it needs no persuasion at all. A message saying that a specific piece of equipment installed in a specific month is now due its inspection is not marketing. It is a service the customer expected to receive and did not, and the response rate reflects that.

The famous statistic, and what it actually says

Any discussion of this eventually cites the numbers: retaining a customer costs five to twenty-five times less than acquiring one, and a five percent lift in retention raises profit by 25 to 95 percent. Both are worth using and both deserve their provenance stated, because they get quoted as if they were laws of physics.

They come from Frederick Reichheld's work at Bain, starting with a 1990 Harvard Business Review article on defection rates in services. The original study looked at credit card companies and insurers, found that a five percent improvement in retention produced more than a 25 percent profit increase, and the wider 25 to 95 percent range came later, in his 1996 book, as an across-industries figure.

A study of credit card companies published in 1990, now quoted at plumbing firms as though the specific multiplier had been measured in their own sector. It was not.

The direction is almost certainly right and holds up across everything I have seen. The specific multiplier is a different matter: it predates the internet, digital acquisition, and every channel a business uses today, and the 5 to 25 range is so wide that it is really a statement that the answer depends entirely on your business. Anybody quoting the top of that range at you has chosen the number that helps them.

Why it never happens

Everyone agrees this is a good idea, which is why the explanation for its absence has to be structural rather than about awareness.

The data is in the wrong system. Customer records live in accounting software, organised by invoice rather than by customer, with no field for what was installed or when it is next due. Extracting a usable list is a genuine afternoon of work, and it is an afternoon nobody has scheduled.

There is also no owner. In a company where everybody is either doing the work or dispatching it, an activity with no deadline and no complaining customer attached simply never reaches the top of anybody's list. It is not deprioritised, it is never prioritised, which is a different and more permanent condition.

And there is a real fear underneath, which deserves acknowledging rather than dismissing. Contacting old customers feels like admitting you need work, and in a business built on referral and reputation that feels like a status loss. The answer is the framing: a maintenance reminder about equipment you installed is a service, and it is received that way.

How to run it without building anything

This is a fortnight of work, most of it once, and then a small recurring habit.

• Export everything from the invoicing system into a spreadsheet, then deduplicate by phone number rather than by name, because the names are inconsistent and the numbers are not
• Tag each record with what was done and, where the work implies a cycle, when the next service is due
• Start with the due list only, in date order, oldest first, and contact them by phone or message rather than by email
• Keep the message specific to what they own and when it was installed, with no offer, no discount, and no marketing language
• Work through it at a rate the diary can absorb, because filling three months of capacity in a week is a customer service problem you created
• Record every outcome back into the list, so the second pass is against real data rather than the same cold export

The pacing item matters more than it sounds. A reactivation list is a stock of demand you can release at whatever speed you choose, which makes it the most useful tool a service business has for filling a quiet month. Releasing all of it at once converts a scheduling asset into a scheduling crisis, and the customers who wait three weeks for the appointment remember that rather than the reminder.

Treated as a dial rather than a pile, it runs backwards from the diary. Decide how many extra jobs a week you actually want, divide by the share who respond, and that is how many people you contact. Nothing else about the list needs deciding.

The response rate is the only figure you have to guess at, and only once: after the first two tranches it stops being a guess, and every row above recalculates itself off your own number. A due-service reminder to somebody whose equipment you installed responds nothing like a cold campaign, so do not borrow a rate from one.

What the table really shows is that the same list is a completely different asset depending on the speed. Run at ten jobs a week it is a two-month burst that leaves you with nothing in reserve and a stretched diary. Run at two it is most of a year of quiet, steady topping-up that nobody in the business even notices, which is usually what an owner actually wanted when they said they should do something with the old customers.

What it becomes once it exists

The first pass through the list is a one-off recovery and it usually produces the largest single number. The more valuable outcome is that the list now exists, is clean, and knows what each customer owns.

From there the maintenance reminders become automatic rather than a project, which converts an irregular pile of past work into something that behaves like a recurring book. That change matters well beyond this year's revenue: contracted or predictable repeat work is the heaviest single input into what a business is worth, so building the reminder system does not only produce cash, it moves the valuation. The mechanics of why are in normalised earnings are a range.

It also changes what happens after a quiet month, which is the practical reason I would do this before any advertising. A business with a worked list responds to a soft quarter by releasing more of it. A business without one responds by spending money on strangers, at the worst possible moment, which is when the cash is tightest. That sequencing argument is the whole of a boring business does not need more leads, and the layer it belongs to is set out in the four-layer growth engine. The full picture is on the system page.

Retention figures originate with Frederick Reichheld's work at Bain & Company, beginning with "Zero Defections: Quality Comes to Services", Harvard Business Review, 1990, and the wider 25 to 95 percent range from The Loyalty Effect (1996). Both predate digital acquisition channels and neither was measured in field-service businesses.