Your 5-Star Reviews Are Hiding a Retention Problem

The Metric You're Celebrating Is Already Old News

My NPS in month two of Reeves Electric was a 4. By month nine, it was an 81. That middle period felt like real progress — the number was moving, reviews were climbing, feedback was good.

What I wasn't tracking: whether any of those satisfied customers were calling back.

NPS and star ratings tell you how a customer felt about the last job. That's useful. It's not the same thing as measuring whether they'll call you again. A 4.8-star average tells you the customers who felt strongly enough to leave a review were happy. That's a narrower signal than most shop owners treat it as — and it hides the actual problem.

I noticed it through attribution, not reviews. CallRail on every channel, job source tracked manually. When I started filtering completed jobs by source and looking at who came back versus who didn't, the reviews looked fine. The return rate did not. Customers were leaving five-star reviews and then, eighteen months later, calling someone else when the next thing broke.

The review is a snapshot of a transaction. The repeat call is evidence of a relationship. Most shops measure one and ignore the other.

---

The Number Almost Nobody Tracks

Pull this: of the residential customers whose jobs you closed eighteen months ago, what percentage have a second job on their record?

Not their reviews. A second job.

Eighteen months is the right window for residential electrical service. It's long enough that the next thing has probably surfaced — an AFCI breaker that starts nuisancing, a GFCI question in a remodeled bathroom, a spouse who finally talked them into the EV charger. Life generates electrical work, or close enough. If past customers aren't calling you for it, they're calling someone else.

At $1.4M across five trucks in 2024, I should be building a repeating customer base year over year. The work is good — the reviews prove that. There's no structural reason those customers need a new electrician every time. When the return rate is low, it usually means nobody followed up. The customer forgot we existed. Not because they were unhappy. Memory is short, and we went quiet after the invoice.

Jobber won't surface this automatically. The data is in there; it won't ask the question for you. I built a query that cross-references job close dates with customer records and flags anyone who's gone 14 months or more without a return visit. First time I ran it, the list was long. That output was more useful than any review report I'd ever pulled — and I'd pulled a lot of review reports feeling good about the numbers.

---

A 4.8-Star Average Doesn't Mean They Remember Your Name

Strong reviews with no reactivation system means you're running a one-time-transaction business. The reviews are real. The satisfaction was real. The relationship wasn't.

The review request is a touchpoint in disguise. Almost nobody treats it that way.

I use NiceJob (Podium works too — pick one, both do the job). Most contractors set it up to chase reviews after job close, watch the stars climb, and stop there. A review request followed by eighteen months of silence is a missed window, not a strategy.

LSA compounds this. It works — I'm not going to pretend it doesn't, we run it — but those leads close at a lower rate because the customer is shopping. They called three electricians; we happened to be fastest. No relationship there. Win the job from LSA, do good work, leave no follow-up system in place, and that customer opens Google again next time the same way they did the first time. You paid for the lead twice without knowing it.

The math on this is straightforward. A repeat customer costs you zero in acquisition. An LSA lead costs real money per booking — and I know what mine costs because I track it through CallRail by channel. Shops that can't tell you their cost-per-booked-job by source are making marketing decisions on instinct. I've been there. It's an expensive place to stay.

---

What the Silence After the Job Actually Costs

When I launched Reeves Electric in February 2022, I had the pricing model, the software stack, the branded vans. What I hadn't built was anything for after the job closed. Invoice sent, review request out, then nothing. Customer went back to their life. We went back to the next call.

The ninety days I spent answering every call myself was mostly about intake — rebuilding the phone screen, the pre-quote questions, the $99 site assessment for complex jobs. But once I was sitting with my dispatcher weekly, listening to calls that booked and calls that didn't, I started noticing a second pattern. The drop-off wasn't during the job. The work was good. The drop-off was happening in the silence after.

We didn't check in when the permit closed. We didn't call at the one-year mark. The customer had a good experience and then had no experience, because we went quiet. When the next thing came up, they didn't think "call Reeves." They thought "I need an electrician" and opened Google, same as before.

A customer who felt fine about the last job but can't recall your name six months later is a lost customer. You'll never see it in the reviews. They didn't leave a bad one. They just left.

That gap — between a satisfied customer and a returning one — is the retention problem. And by the time it shows up in your revenue trend, you've already burned through a year or more of repeat volume you're not getting back.

---

What a Real Reactivation System Actually Looks Like

You don't need an agency for this. You need a trigger, a follow-up task, and a consistent schedule. That's it.

The stack for a five-truck residential shop: Jobber or Service Fusion at the core, Zapier as the connector, NiceJob or Podium for the touchpoint layer. When a job is marked complete in Jobber, Zapier starts a 12-month countdown. At month twelve, a task drops in the dispatcher queue: call this customer. Four minutes. "Hey, this is [name] with Reeves Electric — we did some work for you last year. Just checking in to make sure everything's holding up and see if anything's come up." Recorded through CallRail. Reviewed the same way intake calls get reviewed.

The automated email has its place. The 12-month call is not that place. An email says "we have a sequence." A phone call says "we remember you."

The second piece is a check-in around month six through NiceJob. Not another review request. Something like: "We're six months out from the panel work — wanted to make sure everything's running well and you know we're here." Short. Genuine. That's not marketing, that's service. Customers are usually mildly surprised anyone called. The ones who find it annoying are the exception, not the pattern.

---

What to Do Next Monday Morning

Monday. Open Jobber or Service Fusion. Filter completed residential jobs, closed 15 to 21 months ago. Export the list. For each customer, check whether a second job exists on the record. Count the ones that don't. That ratio is your baseline repeat rate. If you've got 80 customers on the list and 12 have come back, your rate is 15%. Write it down. You can't move a number you've never measured.

The following week. Build one Zapier automation: job marked complete in Jobber triggers a 12-month countdown, which creates a dispatcher task for a personal follow-up call. Test it on two closed jobs before you go live. Route the call through a CallRail tracking number so it's recorded. Pull the recording the first Friday it runs and listen to it.

This Friday. Set a recurring 30-minute block — I do Friday afternoons, same time I do the call recording review with my dispatcher. Who's in the reactivation queue, who got called this week, what happened. It's not a new meeting. It's a new agenda item in something you should already be running.

---

FAQ

My reviews are strong and my customers seem happy — why should I worry about repeat business right now?

Because strong reviews tell you about the last job, not the next one. A satisfied customer who never comes back is invisible in your review dashboard. You won't see the problem until new-customer acquisition costs climb and revenue growth flattens. At that point you've already burned a year or more of repeat volume. Look at this before the number is bad, not after.

How do I pull a repeat customer report out of Jobber or Service Fusion without a developer?

In Jobber, go to the client list, filter by job close date range — 15 to 21 months back — and export. Filter for clients with only one job on record. That's your one-and-done list. Service Fusion has similar filtering in the customer report tab under job history. Neither requires developer work. It's a spreadsheet exercise, maybe 20 minutes, and it shows you something the standard reporting view never will.

Is a 12-month follow-up call too aggressive?

No, if the call leads with service rather than sales. "We wanted to make sure everything's working well" lands differently than "we'd love to earn your next project." Keep it short. Don't try to upsell unless they bring something up. The goal is to exist in their memory when the next thing breaks. Most customers are surprised anyone called, in a good way.

What's a healthy repeat rate for a residential electrical service shop?

I don't have a benchmark for you, and I'd be skeptical of one. What I can tell you is that before we had a reactivation system, our repeat rate was low enough to be embarrassing — I knew this because I'd run the query and looked at it. After running the system for two quarters, it moved meaningfully. Whether your number is "healthy" depends on your market, your average job type, and your mix of large one-time work versus recurring service calls. Know your number. Run the system for two quarters. See if it moves.

Should I automate the reactivation touchpoint or have a real person make the call?

Have a real person make the call. The Zapier trigger and the NiceJob six-month check-in can be automated. The 12-month call should be a human voice — your dispatcher, with the customer name and job details in front of them. Record it through CallRail, review it, and refine the script the same way you'd refine your intake script.

We don't have a dispatcher yet — who makes the reactivation calls?

You do, for now. That's not a bad thing. You'll learn what the conversation actually sounds like, and you'll build the script before you hand it off. When you do hire a dispatcher or CSR, you'll have a real example of what good looks like, not a theory. Start the habit yourself, prove it's worth maintaining, then hire the role that keeps it running. Same sequencing as anything else you've ever hired behind.