Your Repeat Customer Rate Is the Number That Actually Matters

In month two of running Reeves Electric, my NPS was a 4. Intake was broken, communication was broken, follow-through was broken. The low score and the retention problem were the same problem. By month nine, NPS was 81.

Here's what I didn't understand yet: the 81 meant I'd stopped fumbling the experience. It did not mean those customers were calling back.

Those are different things, and I spent most of 2023 learning why.

The Metric You're Celebrating Is the Wrong One

A 5-star review gets written the day the tech cleaned up, thanked the homeowner by name, and turned the lights back on. A return call happens 14 months later when that same customer needs a panel upgrade and has to decide whether to call you or start fresh on Google.

In my shop, those two events weren't connected in the data. The review was in NiceJob. The second job, if it came in through LSA, got logged as a new lead. The connection between satisfied customer and returning customer went untracked.

A shop that generates good reviews is doing solid work and running a decent close-out process. A shop that gets return calls has built something different. You want both. But you should know which one you actually have — and they are not the same report.

What a 5-Star Review Actually Measures

A review captures how the customer felt about one interaction. The tech's demeanor, whether the invoice made sense, whether the problem got solved. What it doesn't capture is whether they'll remember your name 14 months from now or just google "electrician Austin" again.

I know this from CallRail data and a custom Airtable tracking job source against customer history — something I've been running since early 2023. When I cross-referenced the two, I found that a chunk of what I'd been counting as repeat customers weren't organic returns. They'd googled us again and come back through a paid channel. The review didn't retain them. The LSA ad did.

That distinction matters. A customer who gave you 5 stars and then found you again through LSA is still a shopping customer. I've written before that LSA works, it's expensive, and leads close at a lower rate because the customer is comparing at the moment of search. A good review score doesn't change that behavior. The customer who was happy last year will still click the first three results if they can't remember your name. The review tells you they were satisfied. It tells you nothing about whether they came back on purpose.

Why Shops Confuse Satisfaction Data for Retention Data

At a five-truck shop running hundreds of jobs a year, a climbing review score is noise unless you're cross-referencing it against return visit data in your CRM. It feels like signal. It usually isn't.

The review tools make this easy to miss. I use a review management platform — I've recommended both NiceJob and Podium here before, and I'll keep recommending them. They do exactly what they're built to do. Job closes, automated text goes out, customer taps five stars, dashboard looks great. Six months later, that customer calls someone else. The dashboard never knew to ask.

I've said before that most contractors confuse their gut with data they've actually seen. At a five-truck shop in Austin, you're running several hundred jobs a year. The "feels like we're getting a lot of repeat business" read is probably off — maybe high, maybe low — and you won't know which until you pull the number.

The Number You Should Be Watching Instead

Return customer rate. Of all the completed jobs in the last 12 months, what percentage came from a customer with a prior closed invoice in your system?

If you're running Jobber or Service Fusion, this exists right now. Filter completed jobs, last 12 months. Find customers with more than one closed invoice. Divide by total unique customers over the same period. That's your baseline. Four minutes if your records are reasonably clean.

I don't have a benchmark that applies everywhere — residential electrical is different from HVAC, high LSA spend skews you toward net-new, market size matters. When I ran this report for the first time in early 2023, my number was lower than I expected. Meaningfully lower.

Return customer rate belongs in the weekly numbers review alongside booking rate and average ticket. Not monthly. Weekly. If a number isn't on your dashboard and reviewed on a cadence, it isn't being managed. It's just a thing that happens to you.

When My Reviews Were Strong and My Retention Wasn't

In 2023, my Google rating hit 4.9 across 200-plus reviews. I'd put real work into the close-out process — tech communication, invoice clarity, automated review requests. The number was real and it reflected actual effort.

Around the same time, I'd increased LSA spend to chase growth. Revenue went up. New customers came in fast. What I didn't catch right away: return customer revenue as a share of total had drifted down even as the review count climbed. I had new customers coming through the top of the funnel, being satisfied, and then not coming back — not because anything went wrong, but because nobody followed up. I'd built a satisfaction process. I hadn't built a retention process.

The call recordings showed it once I knew to look. I sit with my dispatcher every week and we pull calls — two that booked, two that didn't. When I started sorting by customer history, the conversations sounded different. Return customers had trust already. "We've used you before," then they described the problem. LSA customers were still shopping on the second job even when they'd been happy the first time. They asked about price before they described the problem. The 4.9 rating hadn't changed their behavior at all.

Here's what I actually built to fix it — and it's not a system, it's two things.

A 60-day follow-up sequence through the CRM. Every closed invoice triggers an automation — Zapier off the job-closed trigger in Jobber, routes to a task — that puts a call on my CSR's list 60 days out. Not an email. A phone call. One sentence: "Hey, this is Sarah from Reeves Electric — just making sure everything's still working right." No pitch. That's it. In my shop, roughly one in five of those calls turns into a booked job or a referral.

A dedicated tracking number for customers with prior invoices, routed through CallRail. When that number rings, I know the customer came back looking specifically for us. When a customer with a prior invoice calls the LSA number instead, I know Google did the retention work I should have done. Those two buckets look completely different when you separate them, and the decisions that follow from knowing which one is growing are completely different decisions.

What to Do Monday Morning

Run the report first. Jobber or Service Fusion, completed jobs, last 12 months, customers with more than one closed invoice. Divide by total unique customers. Write that number down. You need a baseline before anything else is worth doing.

Set up the 60-day follow-up next. If you have Zapier or Make, build the automation off the job-closed trigger in Jobber. If you're not there yet on automation, create the tasks manually for now. Start making the calls. One sentence. Your CSR can handle it.

If you're running LSA and haven't split your CallRail tracking numbers by channel and customer history, do that before the end of the week. It's not a long setup. Knowing whether your "returning" customers are actually returning or just finding you through paid search again is one of the more clarifying things you can do to your marketing budget — because the answer changes what you do next.

FAQ

If my reviews are strong and revenue is growing, why does return customer rate matter?

At some LSA spend level — and I've been there — you can have genuine revenue growth while quietly building a business that needs increasing ad spend just to stay flat, because nobody's coming back without a prompt. Strong reviews and growing revenue can coexist with a real retention problem. Return customer rate is how you find out before the ad costs make it obvious.

What's a realistic return customer rate for a residential electrical shop?

I don't have a number that holds across markets and shop sizes, and I'd be inventing it if I gave you one. What I can say is that when I first pulled mine in 2023, it was lower than I wanted, and that was after two-plus years of operation in Austin. The absolute number matters less than the direction — is it moving up or down quarter over quarter? That's the question worth tracking.

How do I tell the difference between a customer who came back on their own versus one who found me through Google again?

Dedicated tracking numbers in CallRail, separated by channel and customer history. Organic return calls one number; LSA traffic routes through a different number. Most shops treat both as repeat business. They're not the same. The organic return is a retention win. The paid return means Google is doing the job your follow-up should be doing, and you're paying for it twice.

My CRM data is a mess going back two years — how do I start tracking this without cleaning everything up first?

Pick a clean start date — first of next month — and tag jobs consistently from there. Pull whatever baseline you can from existing data, even if it's rough. An imperfect baseline is more useful than no baseline. The 90-day trend you'll see going forward will be based on clean data, and that's what you'll actually use. Start from now.

Your Repeat Customer Rate Is the Number That Actually Matters

Enjoyed this article?