Your crew finishes eight jobs in a week, the homeowners are happy, and one of them mentions you were the only company that called back. Then you open the Business Profile and the count still reads eleven reviews, four of them from two summers ago. When you are not getting enough Google reviews, the cause usually sits in one missed moment inside the job: the point where a satisfied customer's experience turns into something the next homeowner can read.
Reviews carry the Trust lever, the one answering "should I believe you?" for somebody comparing three contractors on a phone. They are also the visible output of Customer momentum, the sixth of the growth levers we work through with owners, which asks whether the people you already served would come back or send a neighbor. That momentum feeds Attention right back, since a profile with recent honest reviews gives Google and the homeowner more to work with than a quiet one.
Contractors not getting enough Google reviews are usually running into one of four things
The gap sits in one of four places, and each calls for a different fix.
- Nobody asked, because the ask lives in a technician's memory at the end of a fourteen-hour day.
- Somebody asked, but the customer had to go find you on Google, and names in a small market are easy to mix up.
- The ask went out days late, or through a channel that customer never uses with you.
- The customer did write one, and Google has not published it yet, or has taken it down.
The first three sit entirely on your side. The fourth is the platform, and asking harder does nothing for it.
Count the asks before you change anything
Pull last month's completed jobs and count how many produced an actual review request. A thin count with a healthy ask rate is a different problem from a thin count with almost no asks going out.
Then open your review link on your own phone the way a customer would. It should land on the review composer for the right location, not a general profile page.
What a thin review count does while somebody is comparing
In BrightLocal's 2026 survey of 1,002 US adults, positive reviews made 85% of respondents more likely to use a business, while negative ones made 77% less likely. Those are stated attitudes rather than tracked purchases, so read them as direction and not a formula.
The homeowner hiring you almost never watched you work. They read what your last dozen customers said and decided how much risk they were taking. Housecall Pro's 2025 survey of 1,040 homeowner decision-makers found 73% would refer a professional after an excellent experience, which is the same trust taking a different shape. Reviews just keep working for the people who do not know anybody who has hired you.
Hand people the exact link, not a search
In the current Business Profile you choose Read reviews, then Get more reviews, and copy the link Google generates. That link opens the review composer for your business. Google suggests putting it on receipts, in a thank-you email, or on something you leave behind.
The QR code on that screen generates only in a desktop browser right now, so save it once on a computer for whoever builds your invoices. Custom g.page short names were retired in February 2026, so build the workflow on the generated link. A reviewer needs a Google account, and any email address can create one, which comes up when a customer says they cannot review you.
Ask when the outcome is real, not when the truck pulls away
In GatherUp's 2025 survey of more than a thousand US consumers, 77% wanted a review request within three days. That window opens when the customer knows the work held.
The trigger is a confirmed outcome rather than a job status changing to complete in your software. On an emergency no-heat call, that is the homeowner confirming the house is warm after the technician has driven away. The same logic pushes a roof replacement to the point where inspection and cleanup are accepted, with no driveway full of nails still open, and holds a remodel until the walkthrough and punch list are done.
Those triggers are our operational read, not a Google rule. Google's actual rule prohibits pressuring anyone to leave a review while they are on your premises, and in-home work is close enough that the safe version is the same: mention it on the way out, then send the link after you have gone.
Write it so it sounds like you without steering what they say
A request that works has four parts and no cleverness in it. Say who you are, thank the customer and name the job in your own words, invite an honest account of how it went, and give the link. Keep it optional, since "if you have a minute" does real work in that sentence.
The line you cannot cross is telling the customer what to write. Google prohibits influencing the rating or the content, and its April 2026 wording names two practices directly: asking staff to bring in a certain number of reviews, and asking for reviews containing specific content, including content identifying a staff member. A customer who names your tech on their own is fine. Asking them to is not.
Personalize the wrapper around the request and leave the review itself alone. "Thanks for letting us into the house on a Sunday morning" is yours to write. "Mention Kyle and that we do roofing in Watertown" is not. If the wording is what keeps stalling you, we went through it line by line in how to ask without making the customer feel cornered.
Use the channel the job already used
In GatherUp's question on acceptable request channels, email came in at 53%, face to face at 49%, and text and print at 33% each. Those are stated preferences, not conversion rates, and the published text-versus-email comparisons come from vendors with a stake in the answer.
Send the request down the channel the job already ran on. If every appointment window and arrival notice went by text, the review request belongs in that thread. If the remodel was managed by email with change orders attached, email is where the customer expects to hear from you.
One reminder is reasonable, since forgetting was the most selected reason people gave GatherUp for not reviewing more often, at 42%. A neutral nudge two or three days later respects that. A third and fourth message stops being a reminder.
Make the request part of closing the job
Build the closeout so the request fires whether or not anybody feels motivated that day. Decide which jobs are eligible, what event counts as the trigger, who sends the message, and when the single reminder stops. Log the request date and channel next to the job, so the record lives outside a technician's head.
Two ratios tell you where you stand. Completed jobs to requests sent exposes the missed asks, and it is usually the bigger number to fix. Requests sent to published reviews gives you a conversion rate by channel that beats any benchmark online. Expect some quiet either way, because a fifth of the homeowners Tinuiti surveyed in 2025 never leave public feedback at all.
Send the opportunity to everyone who qualifies, including the customer you suspect is lukewarm. Deciding who sees the link based on how happy you think somebody is turns into review gating, which Google prohibits. Private feedback for service recovery is worth collecting, but it cannot control who gets the public link. The whole workflow is in our review system that runs on job completion.
There is no review count that switches the map on
Google publishes no threshold. Its local ranking guidance says results are based mainly on relevance, distance, and prominence, and that under prominence, more reviews and positive ratings can help.
Consumer comfort is a separate question. BrightLocal found 47% would not use a business with fewer than 20 reviews, which describes homeowner comfort while comparing rather than anything Google does while ranking.
The benchmark worth tracking is the set of companies showing up for your service in your town. Compare their counts, ratings, and how recent their reviews are, and keep the towns separate rather than blending Watertown, Brookings, and Sioux Falls into one target. Recency matters to buyers too, with 74% of BrightLocal respondents looking for reviews written in the previous three months.
Your displayed score averages every published rating, so one new review moves it less the more you already have, and the display can lag by up to two weeks.
The shortcuts that put the profile itself at risk
Gift cards, discounts, prize drawings, and free filter changes tied to a review all violate Google's policy, whatever the ratings say. The FTC's 2024 rule is narrower, targeting incentives conditioned on positive sentiment, so a no-strings drawing can clear that bar and still break Google's rules.
Asking for five stars, handing over wording to copy, or putting a review quota on the crew all shape what gets posted, and Google's April 2026 language names the quota version specifically. A neutral process produces whatever the work earned. An engineered one produces what you told people to say, which is the part being policed.
Google can also do more than delete the reviews in question: for fake engagement it can block new reviews, unpublish what is already there, and put a consumer warning on the listing.
When a review never shows up
Give it several days, since Google checks new reviews and says they can take a few days to appear. If it is still missing, the ordinary causes are a profile merge, a policy removal, or spam detection catching a legitimate review, and Google points owners to support when real reviews disappear. Going the other direction, a review you dislike comes down only if the content violates a policy, because Google removes violations rather than reviews an owner disputes.
This works through repetition, one small habit at the end of every job. The same neutral request goes out whether or not anybody remembers, and over a few months the public record starts matching the quality of the work. That closeout step is what our Reputation Engine runs for owners who would rather not carry it themselves. If you want a second set of eyes on where your requests are going, a free consultation is the low-friction start, and we keep pilots small enough that stopping stays simple.
