A 4.7-star contractor with 180 reviews gets called by every homeowner who searched their trade in their city. A 4.9-star contractor with 12 reviews gets called by almost nobody. The math is not about the average. It is about the count plus the average — Google’s local results, the way homeowners scan, and the way the reviews themselves read all reward volume far more than the average rating once both numbers cross a basic threshold.
This is a piece about building the review system that gets you there, and what to do with the bad review that shows up along the way.
Why reviews compound and most contractors miss it
Reviews behave like compound interest. The first 25 are the hardest to earn, take the most direct asks, and feel awkward. The next 50 come more naturally because the homeowner sees a real review count and feels they’re contributing to a record, not creating one. Past 100, reviews accumulate as a side effect of doing the work — and the contractor who built that base is no longer competing with the contractor who has 12.
The lever most contractors miss: every signed job is a review opportunity, and most are simply not asked. A contractor who closes 200 jobs a year and asks 25 of them for a review will earn maybe 8 to 10 Google reviews — a fine year. The same contractor with the same close rate, who asks every customer at the right moment, earns 60 to 100 reviews. The difference is not in the work. It is in the ask.
The right moment to ask
The window for a review request opens at the end of the job and closes within about 72 hours. Past that, the homeowner’s attention has returned to the rest of their life. The single highest-converting moment is when the lead installer or owner is shaking the homeowner’s hand at the end of the project, the job is paid, and the homeowner is visibly satisfied. The script is short:
“If you have a minute later this week, the single biggest thing that helps us is a Google review. I’ll send you a text with the direct link — you don’t have to write much, even a sentence helps. Sound okay?”
Two details matter. First, the ask comes from the person who did the work, not from a back-office text sent after the truck leaves. Second, it commits to sending the link, which means the homeowner doesn’t have to find the business on Google themselves — a step that loses half the people who said yes.
A polished CRM workflow is helpful but not necessary at low volume. A contractor doing 5 to 15 jobs a week can run this manually: a brief end-of-job script, a templated text from the office phone with the direct review link, and a follow-up text 5 to 7 days later for the ones who didn’t write. At higher volume, tools like Podium, NiceJob, BirdEye, or even just a Zapier flow on top of a basic CRM automate the link send and the follow-up.
Where to send the reviews
Google reviews matter more than every other platform combined for most US trade businesses. The Google Business Profile ratings show up in Local Pack results, drive map-pin click-through, and feed the AI summaries that increasingly answer “best [trade] near me” queries before the user clicks anywhere.
Yelp matters in select markets. The Northeast, parts of California, and pockets of the Pacific Northwest still have homeowners who default to Yelp first. In most of the country, Yelp is a distant second to Google and is best left as a passive presence — a claimed profile, a few reviews, no paid placement.
Trade-specific platforms (Angi, HomeAdvisor, Houzz, Thumbtack) are a different category. They serve dual purposes: as review repositories and as paid lead sources. The lead-source side is a separate conversation; the review side is fine but should not be the primary place reviews accumulate. Reviews on a platform you don’t control can be reorganized, paywalled, or downranked at the platform’s discretion. Google reviews show up everywhere Google shows up.
A simple rule: ask every customer for a Google review by default. If they mention they came from Yelp or Angi, mention that platform too. Otherwise stay focused.
The reviews that actually get written
A homeowner who agrees to write a review and then doesn’t follow through is the most common outcome. Two interventions raise the follow-through rate substantially.
The first is the direct link to the review form, not the Google Business Profile page. The shortcut is to find your business on Google, click “Write a review,” copy the URL, and use that link in every ask. It bypasses three clicks that lose people.
The second is a brief prompt rather than an open page. “Even a sentence about what we did and how it went is useful — no need to write a paragraph.” Homeowners stall on writing reviews because they think the format is a long-form opinion. Telling them otherwise unlocks the ones who would have closed the tab.
Two follow-up texts is the upper limit. One at 5 to 7 days, one at 14 days. Past that, the ask costs goodwill faster than it earns reviews. A homeowner who hasn’t written by week three was never going to.
What to do with a bad review
A bad review will land at some point. It is not the end of the world, and it is not a failure of business operations. It is a moment to demonstrate, on a public page, how the business handles complaints. The bad review and the response together are read by more prospective customers than ten good reviews in a row.
The response template that works on almost every bad review:
“[Customer name], thank you for taking the time to write this. I’m sorry the [specific issue they raised] didn’t go the way it should have. We take this seriously and would like to make it right. I’ve left you a voicemail and sent a text — please reach out at [phone] so we can work through this. — [Name], Owner”
What that does: acknowledges the customer by name (signals you’re not a chatbot), names the specific issue (signals you read the review), apologizes without admitting legal fault, commits to making it right, and gives a real way to reach the owner. Length: 4 to 6 sentences, max. A long response reads as defensive even when the content isn’t.
Three things to avoid in a public response: arguing the facts on the public page, blaming the customer for any part of what happened, and the corporate-counsel non-apology (“we regret any inconvenience caused”). All three convert a single bad review into evidence the business is hard to work with.
The actual resolution happens off the public page. A phone call, an offer to revisit the job, sometimes a partial refund, sometimes a fix at no cost. If the customer updates the review afterward — many do, when handled well — even better. If they don’t, the response itself still does most of the work for future readers.
Handling fake or extortion reviews
A small percentage of reviews are not legitimate complaints. Examples: reviews from people who never used the business, reviews demanding free service in exchange for taking down a negative star rating, competitor reviews from sockpuppet accounts. Google’s policies prohibit these and there is a flag-for-review process inside the Google Business Profile dashboard.
Two practical notes. First, fake reviews flagged with documentation (screenshots, a written description of why it violates policy) are removed more often than flagged-only reviews — Google’s review team responds to context, not just clicks. Second, an extortion attempt should be reported to Google and, if the dollar amount is meaningful, to local law enforcement. Paying for review removal is a one-way ticket to repeat visits from the same attacker.
The reviews that aren’t removable but are nakedly unfair (a one-star review for not answering a Sunday call, say) are handled with the response template above. The unfairness shows in the response juxtaposition; future readers see it without being told to.
The numbers worth tracking
Four metrics are worth tracking on the review program.
- New reviews per month, by platform. Trend matters more than absolute count.
- Conversion rate from ask to review. A contractor asking 60 customers and earning 18 reviews is at 30%. Below 15% suggests the ask timing or the link is failing. Above 40% suggests the system is working.
- Average rating, last 90 days. A drifting average is an early signal of a job-quality or crew-management problem.
- Response rate to negative reviews. A response to every negative review within 48 hours is the standard. Mixed response rates send a worse signal than no responses at all.
These don’t require a dashboard. A monthly note in a spreadsheet works.
What this looks like at scale
A contractor doing 200 jobs a year with a working review system earns 60 to 100 reviews annually after the first year. By year three, the cumulative count is at 180 to 300, an average around 4.7 to 4.9, and the Google search results are no longer a question — every “best [trade] in [city]” search in the local area surfaces them at or near the top of the Local Pack. The marketing budget that used to go into paid ads can be reallocated, because the reviews are doing the work the ads used to.
This is slow growth. It is also free, durable, and the kind of asset that doesn’t disappear when an algorithm shifts.
The cheapest growth available, again
This is the second piece in a small series about the work most contractors don’t do well — the parts of the business outside the trade itself. Lead response is one. Reviews are another. Both cost almost nothing to fix and both raise the close rate on every lead the business already pays for.
At Reliable.Work, every inbound homeowner inquiry from a city we cover routes to a single verified contractor for that metro — no shared queue, no auction, no race to the bottom on price. If you are a contractor who wants to be the only one in the running when a homeowner in your territory hits send, the partnership application is at /for-pros/. Reliable.Work is a California lead referral service, not a contractor; the work belongs to the partner who owns the territory.