Customer retention: how many customers come back?

Why retention is the number to watch

Every home service company knows its new-lead count. Far fewer know how many of last year's customers came back. Yet repeat customers are usually cheaper to serve: no ad spend to win them, fewer surprises on site, and a higher chance of saying yes to a recommendation from a technician they already trust.

Demand keeps coming. The average US household completed 10 home projects in 2025, up from 9^[1], and regular maintenance is the most common planned project, ahead of interior painting^[1]. Homeowners are also short of options: on average 53% of surveyed homeowners reported difficulty finding pros in 2025^[1]. A customer who had a good experience with you has every reason to call you again, if you stay in touch.

The basic formula

The standard retention formula compares the customers you kept with the customers you started with:

Example: a fictional pest control company starts the year with 1,200 active recurring customers, wins 400 new ones during the year, and ends with 1,300 active. Retention = (1,300 − 400) ÷ 1,200 = 75%. A quarter of the starting customers left during the year, and new sales covered the gap.

The formula works when ‘customer’ has a clear status, as it does for recurring services. For trades where customers call when something breaks, you need a different definition.

Pick the version that fits your trade

Repeat-purchase rate is simple: of the customers who paid you in a given period, how many paid you again within the next 12 or 24 months? Example: a fictional electrical contractor served 800 households in 2024; 200 of them booked again by the end of 2025. Its 12–24-month repeat rate is 25%.

Measure by cohort, not by total

A single company-wide percentage hides what is going on. Group customers by when and how they first came to you, then follow each group.

1. 01

   Export customers with their first job date

   From your job management system: customer ID, first paid job date, lead source, trade and every later paid job date.

2. 02

   Group by first-job quarter

   Each quarter's new customers form a cohort.

3. 03

   Count who came back

   For each cohort, count customers with a second paid job within 3, 6, 12 and 24 months.

4. 04

   Split by source and trade

   Compare customers from referrals, search, lead platforms and maintenance-plan sign-ups.

5. 05

   Compare like with like

   This year's spring cohort with last year's spring cohort, never spring with winter.

Cohorts often show that customers from one lead source rarely return, while customers who joined a maintenance plan on their first visit return at a much higher rate. That is the information that changes marketing budgets.

What drives the number

Most of your customers own the homes you work on: the US homeownership rate was 65.5% in the third quarter of 2026^[2]. Owners keep the same furnace, roof and lawn for years, so the relationship can last as long as they stay in the house. These are the levers that keep them calling you.

• Maintenance plans and recurring schedules. The single strongest structural driver, because the next visit is booked before the customer has to think about it.
• Reminders that arrive at the right time: before the cooling season, before the first frost, at the quarterly pest visit.
• Callbacks handled well. A problem fixed quickly and without argument often creates a more loyal customer than a visit that went perfectly.
• The same technician or crew where possible. Customers return to people, not logos.
• Clear prices and invoices that match the estimate.
• Reviews and reputation, because returning customers check them too: 82% of US adults at least sometimes read reviews before a first purchase^[3].
• Things you can't control, such as customers moving. Remove moved-away customers from the denominator if you can identify them.

Find out why customers leave

Churn reasons are easy to collect for recurring services and hard for everything else. For cleaning, lawn and pest plans, ask every canceling customer one question and record the answer in a fixed list: moved, price, quality, scheduling, switched to another company, no longer needed, other. For irregular trades, look for the customers who used you once and then booked a competitor for the next job; a short survey to a sample of lapsed customers is often enough to see the pattern.

HVAC companies have a large base to work with: 88% of US households use air conditioning^[4]. Most of those households will need maintenance or a repair in the next few years. The question is only whether they call you or someone else.

When a reason appears again and again, assign it an owner. Scheduling complaints belong to dispatch, quality complaints to the service manager, price complaints to the owner. A named person with a monthly number to move will find fixes that a general resolution to ‘improve service’ never will.

A one-page monthly retention report

Keep the report small enough that the owner reads it every month. Five lines are enough, each compared with the same month last year.

1. Active customers at the start and end of the month, using your fixed definition.
2. New customers won, and how many of them joined a maintenance plan or recurring schedule on the first visit.
3. Customers lost, with the top three reasons from your cancellation list.
4. Repeat-purchase rate for the cohort that first bought 12 months ago.
5. Callbacks and complaints, with how many of those customers have booked again since.

Add one sentence on what you will change next month. A report that never leads to a change becomes a ritual, and the team stops looking at it.

Mistakes that distort the rate

• Counting every name in the database as a customer, including leads who never paid.
• Changing the definition of active halfway through the year.
• Mixing commercial accounts with homeowners; their patterns are completely different.
• Ignoring duplicates created when the same household books under two phone numbers or two spouses' names.
• Comparing a busy season with a quiet one and calling the difference a trend.

Benchmarks: be careful

You will see retention ‘benchmarks’ for home services quoted online, often without a named study, sample or definition. There is no reliable public dataset of retention by trade. A number that mixes weekly cleaning with once-a-decade roofing tells you nothing. Your own history is the benchmark: measure the same way every year and work to beat it.

Example: one cohort tells the story

Example: a fictional HVAC company split its 2024 new customers by source. Customers who came through a lead platform booked again within 12 months at a much lower rate than customers referred by neighbors, and very few joined a maintenance plan. The company didn't stop buying leads, but it added a maintenance-plan offer to every first visit and a reminder call before the next season. A year later, the lead-platform cohort's repeat rate had moved closer to the referral cohort.

Turn retention into money

Retention matters because of what each returning customer is worth. Use your rate to calculate customer lifetime value, then decide how much you can afford to spend to win a new customer. And look hard at callbacks: turning a callback into a loyal customer is one of the fastest ways to lift the number.