Your Post-Job Follow-Up Is Too Late and Too Generic

Month two of Reeves Electric, my NPS was a 4. Not 40. Four.

I had done everything right on paper. Branded vans, flat-rate pricing, Jobber configured before we turned the first wrench. What I had not built was anything that happened after the invoice got paid. Good electrical work, happy customer, natural repeat — that was the whole plan. That plan cost me a year of goodwill I had to rebuild from scratch.

The 90-day intake rebuild is the story I tell most often — answered every call myself, recorded the patterns, restructured the process, NPS went from 4 to 81 by month nine. What I don't talk about as much is what I found when I started looking at what happened after the tech left the driveway. Intake I knew was broken because I was watching it break in real time. Post-job, I thought the automated review request was a strategy.

It wasn't.

The 48-Hour Window

Residential customers make their repeat-and-refer decision in the 48 hours after the invoice clears. Not at the one-year mark when your drip email lands. Not when they see your truck in the neighborhood next spring. Right now, tonight, when they're deciding whether to mention you to their neighbor or just move on.

If you do nothing in that window, you don't get a neutral outcome. The customer liked you fine, had no complaint worth voicing, and never thinks of you again because you gave them nothing to anchor to. No referral. No callback next year when the panel starts acting up.

Most shops do nothing in that window except wait. Maybe the Jobber review request fires at 24 hours. Maybe the customer fills it out. The owner checks the review dashboard Friday, sees 4.8, figures things are fine, and moves on. That's the entire post-job process — an automated request and a hope.

What the Review Request Actually Does

I run Jobber at the core and NiceJob for review management. That stack works. The mistake I made early — and I've watched other shops make it — is treating the automated review request as retention work.

It isn't. A review request after close serves one purpose: it gets you reviews. Reviews matter. They affect your LSA ranking, your Google profile conversion, the call you get from a customer who spent 20 minutes vetting you before dialing. I'm not dismissing it. But it's the last step in a sequence, and most shops are running it as the only step.

When my dispatcher and I started pulling call recordings weekly, we caught something that should have been obvious. Techs were closing out jobs without setting any expectation for what came next. The customer got great work — panel cleaner than they'd ever seen it — and the tech was in the driveway, truck running, zero mention of the AFCI breaker that was borderline, zero mention of what to watch for on that 40-year-old subpanel, nothing about when to call us back or why.

The customer had no idea what they didn't know. We gave them nothing to hold onto.

The one-year drip sequence has the same problem, just slower. I'm not saying kill it. But if the first 48 hours went badly, or were just empty, no drip sequence saves it. You're emailing a stranger. Fix the 48 hours first. Build the drip second.

The Actual Sequence

Here's what I run, in order.

Same day, before the tech leaves the driveway. Closeout conversation at the door. Scripted. Consistent across every tech. The script is in the next section.

24 hours after close. A personal outreach from the office — not automated, not a merge field with my name in it. A real person who references the actual job. Not "Hi Karen, thanks for choosing Reeves Electric!" but "Hey, wanted to check in on the panel work from yesterday — did everything look good when you tested the new circuit?" Two sentences. Specific. Human.

The review ask goes here, tied to that 24-hour contact. When the CSR gets a positive response to the check-in, they send the review link. The customer is already thinking about the job. The ask lands in context instead of in the void.

48 hours after close. A second touch that has nothing to do with reviews. Did anything feel off? Is there anything we didn't explain well? Do you have questions about what we found? This call runs about 40% answer rate for us. The people who answer almost never have complaints. But the call itself signals we're not done with them just because the invoice cleared.

The Tech Closeout

I have a senior tech who's been with me since the Lonestar Electrical Services days. Best technical guy on the crew. Zero callbacks for warranty. Customers who watch him work love him.

His reviews were averaging 3.6 stars. I couldn't figure out why until I pulled the recordings.

The pattern was right there. His review issues clustered on days he was running behind — late-week, after a long job went sideways. On those days, the closeout was nothing. He'd hand the customer a receipt on his tablet, say "all good, any questions?" and be in the truck before they processed the answer. The reviews said "didn't really explain what he did" and "felt a little rushed at the end." Both accurate. Neither was about the quality of the actual work.

His teammates ran the same compressed schedule. Their reviews were 4.8. They naturally did the soft close, the verbal walkthrough, the "here's what to watch for." He didn't. Nobody had ever scripted it because everyone assumed it was obvious.

At the end of a long day with the next job 40 minutes across town, nothing is obvious.

What I built was a four-line closeout script. Every tech, every job, before they leave the driveway.

1. What we did today
2. What we found that we didn't fix, and why
3. What to watch for
4. What to do if something feels off

The senior tech's reviews went from 3.6 to 4.7 in one quarter. I didn't change what he did on the job. I changed what he said when he left.

Attribution Applies Here Too

I track marketing attribution obsessively — CallRail numbers on every channel, UTM tags, a custom Airtable that logs source, job value, and booking rate by channel. What took me longer was applying the same logic to retention.

Most shops have a vague sense that "referrals are good for us" or "we get a lot of repeat customers." That's the same as saying "most of our work comes from Google." Might be true. You don't actually know.

In Airtable, I tag repeat-customer calls differently from new-customer calls at intake. My CSR flags it — takes five seconds. At the end of each quarter I can pull repeat-contact rate by quarter, by job type, and by technician. That last cut is where the data gets uncomfortable in the best way.

The specific number I watch: percentage of customers who had a service call in the last 18 months and called back for a second job within that window. I track it monthly, right alongside revenue. Not because it's the most important number in the business — it isn't — but because it's the number that tells me whether the 48-hour process is actually working or whether I'm just in a good market with good timing.

If you're growing and you've never measured repeat-contact rate by tech, you have at least one tech who is quietly exiting customers out of your pipeline. The revenue loss doesn't show up anywhere. It just never shows up.

Monday Morning

Pull your last 20 closed invoices from Jobber. Check the timestamps on every outbound contact after close. What fired, and when? If the only thing that went out was the automated review request, that's your baseline. Write it down.

Write the four-line closeout script this week. What we did, what we found but didn't fix and why, what to watch for, what to do if something feels off. Keep it to four lines — techs won't use a paragraph. Walk through it at the next weekly huddle, actually roleplay it once, then pull three post-job check-in recordings before Friday. The CallRail recordings will tell you the truth faster than asking.

Tag your next 30 incoming calls as new or repeat at intake. One field in Jobber or a note in your dispatch sheet. Thirty days from now you'll have a real repeat-contact rate for the first time. It won't be perfect. It'll be real.

---

FAQ

My Jobber review request already goes out automatically — isn't that handling it?

It's handling one piece of it. The automated review request gets you reviews, and reviews matter for your Google presence and your LSA conversion rate. But it fires into the void with no context, no human voice, no reference to the actual job. Keep it, but don't confuse it for a post-job process. It's one step in a sequence that needs three or four more steps around it to do actual retention work.

We send a follow-up email after every job. What are we probably missing?

Specificity. Most follow-up emails are merge-field templates — "Thanks for choosing [Company Name]" plus a link. The customer has no reason to engage because nothing in it refers to their actual job. One sentence about what was done specifically. One sentence about anything noted but deferred — the borderline breaker, the old receptacle. A direct line to reach someone if anything feels off. That's the whole thing.

How do I get my techs to do a closeout script without it sounding robotic?

Script the content, not the delivery. The tech who naturally says "so here's what we got done today" will say that. The one who's more direct will deliver it differently. What matters is that all four points land, not that they sound identical. Roleplay it once at a huddle — five minutes of practice removes most of the resistance. Then pull recordings for the first two weeks and give individual feedback. Takes about three weeks to feel normal.

Is a 48-hour follow-up call realistic when we're running five trucks and the owner is still partly in the field?

Yes, but it's not the owner's call to make. At five trucks you have the volume to justify a part-time CSR. If you don't have one yet, that's the first hire — not because of the follow-up call specifically, but because you're spending owner-rate hours on $25-an-hour work. I've written about that math before. The follow-up call takes two minutes per job. A CSR at 20 hours a week handles it alongside dispatch and intake with capacity left over.

What repeat-customer rate should I expect for residential electrical service?

I can only tell you what I see in my own shop. I don't have a clean industry benchmark. In my numbers, somewhere around 40% of customers from a given 18-month window come back for a second job within that window. Whether that's high or low for your market, I genuinely don't know — your customer age, home age, and local competition all move it. The more useful question isn't what the number is. It's whether the number is moving. Track it for two quarters and see the direction.