You just wrapped a furnace replacement. Customer’s thrilled. House is warm, invoice is paid, crew is packing up. You’re already thinking about the next job. Three days later you remember you meant to ask for a review. You text the customer. They say sure, they’ll get to it. They don’t.

This happens in every trades business. The work is great, the customer is happy, and the review never shows up. It’s not malice. It’s friction. The customer has to remember, find the link, open Google, write something. By day four they’ve moved on.

The math is straightforward. If you close 200 jobs a year and 40% of happy customers would leave a review when asked at the right moment, that’s 80 reviews. If your manual follow-up captures 20 of them, you’re leaving 60 on the table. In a local service business, 60 reviews over twelve months is the difference between page one and page three. It’s the difference between your phone ringing and your competitor’s phone ringing.

This isn’t a marketing problem. It’s an operations problem. You can’t ask for a review while you’re torching a pipe or pulling wire. Your dispatcher can’t send review requests between routing three crews and taking emergency calls. The admin who cuts invoices doesn’t have a system that says “this job closed yesterday, send the review ask now.”

So it doesn’t happen. Or it happens inconsistently. And you leak $50,000 to $200,000 a year in jobs that would have come from better visibility.

What Review Collection Actually Looks Like in a Trades Business

Let’s walk through the current state. Job closes on Tuesday afternoon. Plumber finishes a slab leak repair, customer signs off, payment clears. The plumber texts the office “done, customer happy.” That’s the last structured touchpoint.

If you’re organized, the invoice goes out that night with a thank-you note. If you’re very organized, someone manually adds a task to follow up in two days. If you’re running on adrenaline and a whiteboard, the job just closes and everyone moves to the next emergency.

The review ask, if it happens at all, comes a week later when you’re looking at your Google Business Profile and realize you haven’t had a new review in a month. You scroll through recent invoices, pick three customers, and send a text. One replies “already did it” (they didn’t). One says “sure, send me the link again.” One ghosts you.

This is the manual tax. It’s not that you don’t care about reviews. It’s that review collection requires the same timeline discipline as follow-up on estimates, and you’re already underwater on that. The work that drives revenue today always beats the work that drives revenue in six months.

The second problem is timing. Customers are most willing to leave a review within 24 hours of a great experience. After three days, willingness drops by half. After a week, you’re asking them to remember details and reconstruct enthusiasm. It feels like homework.

The third problem is channel. If you text a Google review link, half your customers will open it on their phone, get lost in the Google login flow, and bail out. If you email it, it sits in their inbox behind 40 other emails. If you ask in person at the end of the job, they’ll say yes and forget by the time they’re back inside.

What works is a message sent within 24 hours, with a direct link, from a system that knows the job is closed and the customer is happy. And then a second nudge 48 hours later if they haven’t done it. That’s not a human task. It’s an agent task.

How the Review and Reactivation Agent Works

The Review and Reactivation Agent sits downstream of your dispatch or invoicing system. When a job status flips to complete, or when an invoice is marked paid, the agent gets a trigger. It waits six hours (enough time for the crew to leave and the customer to settle in), then sends the first message.

The message is short. “Hey [Name], thanks for trusting us with your water heater replacement yesterday. If you’ve got 60 seconds, we’d love a quick review. Here’s the link: [direct Google review URL]. – [Business Name]”

The link goes straight to the review form, pre-filled with your business name. No login maze, no search. One tap, they’re writing.

If the customer leaves a review, the agent logs it and stops. If they don’t, the agent sends a second message 48 hours later. “Hi [Name], just wanted to make sure you saw this. We really appreciate your feedback. [Link].” If they still don’t, the agent waits two weeks and sends a final soft ask. Then it stops. No one wants to be nagged into a review.

The agent also tracks sentiment. If the job notes say “customer complained about timing” or “had to come back to fix a leak,” the agent doesn’t ask for a review. It flags the job for a personal follow-up call instead. You don’t want a 2-star review because a bot asked someone who was already annoyed.

This is where the reactivation piece comes in. The same agent that asks for reviews also tracks service intervals. If you installed a furnace, the agent sets a reminder for eleven months. If you did a seasonal AC tune-up, it sets a reminder for next spring. When the interval hits, the agent sends a reactivation message. “Hey [Name], it’s been a year since we installed your furnace. Want to get on the schedule for a tune-up before winter hits? Reply YES and we’ll call you to book it.”

That message converts at 15% to 25% depending on the trade and the relationship. It’s not cold outreach. It’s a timely nudge to someone who already knows and trusts you. And it happens automatically, whether you’re on the tools or on vacation.

The agent doesn’t replace your CRM. It plugs into whatever you’re already using (ServiceTitan, Housecall Pro, Jobber, or even a Google Sheet if that’s where your jobs live). It reads job status, sends messages, logs responses, and hands off to a human when the customer replies with a question or wants to book.

We build this as part of the AI audit for trades businesses. The audit maps your current workflow, identifies the handoff points where reviews and reactivations fall through, and specs the agent to fit your dispatch cadence and customer communication style. It’s not a generic bot. It’s tuned to how your business actually closes jobs.

The Dollar Reality of Missed Reviews

Let’s put a number on it. You run an HVAC company doing $3 million a year. You close 250 jobs. Average ticket is $12,000 (mix of service, repair, and replacement). You’re good at what you do. Customers are happy. You have 40 Google reviews.

Your competitor down the road does $2.5 million. Similar quality. They have 120 reviews. When someone searches “HVAC repair near me,” your competitor shows up first. Their phone rings 30% more often than yours. They convert at the same rate you do, but they get more at-bats.

If 30% more inbound leads means 15 more jobs a year at your average ticket, that’s $180,000. You’re not losing that revenue because your work is worse. You’re losing it because your follow-up system is a human trying to remember to send a text.

Now assume you implement the Review and Reactivation Agent. It asks every customer for a review within 24 hours. It nudges twice if they don’t respond. It skips the customers who had a problem. Over twelve months, you go from 40 reviews to 95 reviews. Your average rating stays the same (you’re still doing the same quality work), but your volume crosses the threshold where Google starts showing you in the top three local results.

Inbound calls go up 20%. You close 50 more leads. At a 30% close rate, that’s 15 jobs. At $12,000 average, that’s $180,000. The agent cost you $4,000 to build and $200 a month to run. You’re up $170,000 in year one.

The reactivation side is harder to model because it depends on your service mix, but the ranges are consistent. If you have 500 past customers and you reactivate 20% of them over two years, that’s 100 jobs you wouldn’t have gotten otherwise. Even if half of those are small service calls at $800, you’re looking at $40,000 in recovered revenue.

This isn’t theory. We see it in every trades business we audit. The owners who grow fastest aren’t the ones with the best trucks or the flashiest website. They’re the ones who close the loop on every job. They ask for the review, they follow up on the estimate, they reactivate the customer when it’s time for the next service. And they do it with systems, not heroics.

If you want to see what this looks like in your business, book a 60-min Omni Audit. We’ll map your current close-out process, calculate your review gap, and spec the agent. You’ll walk out with a workflow diagram, a cost model, and a build plan. No deck, no discovery phase, just the three outputs you need to decide.

What Happens When You Combine Review Collection with Dispatch

The Review and Reactivation Agent works on its own, but it’s more valuable when it’s part of a wider automation layer. Most trades businesses we work with start with two agents: the 24/7 Dispatch Voice Agent and the Review and Reactivation Agent.

Here’s why they stack well. The Dispatch Voice Agent answers every call, qualifies the job, and books it into your schedule. It captures the customer’s name, phone, address, and job type. That data flows into your dispatch system. When the job closes, the Review and Reactivation Agent picks it up and sends the review ask.

The handoff is clean because both agents are reading from the same system. There’s no manual data entry, no CSV export, no “remember to add them to the review list.” The customer calls, the job gets booked, the crew does the work, the review ask goes out. It’s one workflow.

The Dispatch Voice Agent also solves the missed-call problem that costs most trades businesses $500 to $3,000 per lost job. If you’re running two crews and taking 40 calls a week, you’re missing 10 of them. Half don’t leave a voicemail. That’s $10,000 to $60,000 a year in jobs that never made it onto the schedule.

When you add the Estimate Follow-Up Agent to the mix, you’ve automated the three highest-value follow-up tasks in a trades business: answering the phone, closing the estimate, and asking for the review. Each one is worth $50,000 to $150,000 a year depending on your volume. Together, they’re the difference between a business that grows 10% a year and a business that grows 30%.

We cover the full stack in our guides on AI for trades operations, but the review agent is the easiest place to start because it doesn’t touch your dispatch workflow. It sits downstream, watches for closed jobs, and sends messages. Low risk, high return, and you’ll see results in 30 days.

The Follow-Up Discipline That Separates Growing Shops from Stuck Shops

Every trades business has a follow-up problem. Estimates go out and sit. Customers say “we’ll think about it” and you never hear back. Jobs close and no one asks for a review. Past customers fall off your radar until they call a competitor.

The shops that grow are the ones that follow up. They call the estimate on day two. They text on day five. They send a final nudge on day fourteen. They ask every happy customer for a review. They reactivate every past customer at the right interval. It’s not complicated. It’s just consistent.

The problem is that consistency requires either a dedicated admin (expensive, hard to hire, burns out) or a system. Most owners try to do it themselves. They set reminders, they batch it on Friday afternoons, they feel guilty when they don’t get to it. It works for a month, then a big job comes in and follow-up falls apart.

This is where AI agents earn their keep. They don’t get busy. They don’t forget. They don’t take vacation. They just execute the workflow you defined, every time, on schedule. The Review and Reactivation Agent sends the review ask 24 hours after every job. The Estimate Follow-Up Agent texts every open estimate on day two, day five, and day fourteen. The Dispatch Voice Agent answers every call, even at 9 PM on a Sunday.

You get the consistency without the overhead. And because the agents log everything, you can see what’s working. If review requests sent within 12 hours convert at 50% and requests sent after three days convert at 20%, you adjust the timing. If reactivation messages sent in October (before winter) convert better than messages sent in January, you shift the schedule. You’re not guessing. You’re tuning a system based on your own data.

We’ve also put together a practical worksheet that walks through the after-hours and follow-up gaps most trades businesses have. It’s called the After-Hours Call Recovery Plan for Trades, and it’ll help you map where calls, estimates, and reviews are falling through right now. You can grab it here: After-Hours Call Recovery Plan. It’s a 20-minute exercise, and it’ll give you a clear picture of your current leakage before you talk to anyone about automation.

What the Audit Looks Like

The Omni Audit is 60 minutes. We start with your current close-out process. How do you know a job is done? Who sends the invoice? Where does job status live? How do you currently ask for reviews, if at all?

Then we map the gaps. How many jobs closed last month? How many of those customers got asked for a review? How many left one? What’s your current review velocity, and what would it need to be to move up in local search?

Then we spec the agent. What triggers it? What does the first message say? What’s the timing on the follow-up? What happens if the customer had a problem? Where does the agent log the response?

You walk out with three things: a workflow diagram that shows the current state and the future state, a cost model that breaks down build cost and run cost, and a build plan with a timeline. If you want to move forward, we start the build the next week. If you don’t, you keep the diagram and the cost model and you can hand it to someone else or build it in-house.

No deck. No discovery phase. No “we’ll get back to you with a proposal.” You get the outputs in the room, and you decide.

Book your Omni Audit here. If you’re doing over $1 million in revenue and you’re not capturing 50% of your potential reviews, this is $150,000 sitting on the table. Let’s go get it.