Mistake #1: Ignoring the bad review instead of responding to it
A one-star review sits on a contractor's Google profile for eleven months with no owner response. That's the single most common thing we find when we pull up a new prospect's profile. The homeowner scrolls past four-star jobs, reads the one 1-star at the top (Google often surfaces recent or "most relevant" reviews near the top regardless of star count), sees no reply, and assumes the worst: either the complaint is true, or the business doesn't care enough to answer. Either read costs the click.
Responding does three things a lot of contractors don't expect. First, it shows the next reader you're not hiding from criticism. Second, a calm factual response often walks back the reader's assumption even when the review itself stays up (Google rarely removes reviews just because they're unflattering; removal requires a policy violation, not just an unhappy customer). Third, and this matters more every quarter: AI answer engines pull review text and owner responses as source material when a homeowner asks something like "who's a reliable plumber near me." An unanswered complaint is a citation with only one side of the story in it.
The fix isn't complicated, but it has to happen on a schedule, not "whenever someone remembers to check." Every review, good or bad, gets a response within 48 hours. Bad reviews get: acknowledge the specific issue by name (not a copy-paste apology), state what was done or offered to fix it, and invite the person to call directly. No arguing in public. No admitting fault you don't owe. Just calm, specific, and human.
- Unanswered 1-star reviews read as unaddressed problems, whether or not they are
- Google review removal requires a policy violation (fake, off-topic, harassment), not just a bad experience
- Response text becomes part of what AI answer engines cite when summarizing a business
- A 48-hour response window is realistic for an owner-operator; longer than a week reads as absent
Mistake #2: Asking every customer the same generic way (or not asking at all)
"Please leave us a review" texted to every customer the same day the invoice goes out is barely better than not asking. Response rates on a generic blast run low because the ask has no context and no urgency attached to a specific good experience. The customers most likely to actually open the link and write something are the ones asked at the right moment, with the right framing, not the ones added to a mass list.
The other half of this mistake is simpler: a lot of established contractors just don't ask at all. They assume good work speaks for itself. It doesn't, on Google. A homeowner who had a fine experience and wasn't asked will not go write a review unprompted nine times out of ten. The reviews that show up organically without an ask skew toward the extremes: someone furious, or someone unusually delighted. That's how a contractor with hundreds of satisfied jobs a year ends up with twenty total reviews and a rating dragged down by the two loudest outliers.
What actually moves the needle: timing the ask to the moment the customer is happiest (immediately after final walkthrough, not two weeks later after an invoice dispute), personalizing the message with the crew member's name and the job type, and using a direct link that drops the customer straight onto the review form instead of making them search for the business first. Text performs better than email for most home-service trades because it gets read within minutes, not buried in a promo folder.
| Ask method | Typical response behavior |
|---|---|
| Mass email blast, no personalization | Low open rates; reviews skew to extremes |
| Generic text, sent days later | Moment has passed; low completion |
| Personalized text at walkthrough, direct link | Highest completion; captures the satisfied middle, not just extremes |
Mistake #3: Review-gating (and not knowing where the legal line sits)
Review-gating is filtering customers before they leave a review, usually by asking "how was your experience?" first and only sending happy respondents on to Google while quietly not following up with anyone who says "okay" or worse. Google's review policies prohibit this specifically: you cannot selectively solicit reviews based on the customer's likely sentiment, and profiles caught doing it can face manual actions, from review removal to reduced visibility in the map pack.
Contractors get here honestly. A well-meaning office manager builds a filter step because nobody wants to funnel an angry customer straight to a public review form. The instinct is reasonable. The mechanism is against policy. The line that matters: you can ask everyone the same way and let outcomes fall where they fall, and you can have a separate internal process for catching problems before they escalate (a post-job check-in call, a satisfaction survey that's clearly internal and not gated to reviews). What you cannot do is route only the likely five-star customers toward the review link.
The safer, and frankly more effective, version of this same instinct: ask every customer, every time, through the same link, and build a fast internal response process for when a bad one lands (see Mistake #1). Reserve any pre-screening only for internal quality tracking that's clearly separate from the review ask, never as a gate in front of it.
- Google prohibits selectively soliciting reviews based on anticipated sentiment
- Enforcement can mean review removal or reduced profile visibility, not just a warning
- Ask everyone through the same link; use internal follow-up calls for quality control, not gating
- A fast public response beats a hidden filter for managing the occasional bad review
Mistake #4: Waiting too long after the job to ask
The ask that goes out three weeks after completion competes with a homeowner's fading memory, a new project they've already moved on to, and a phone full of other notifications since the crew left. The window where a homeowner is actually thinking about the work, warm on the experience, and likely to spend ninety seconds writing something specific, is short: same day to about 48 hours after the final walkthrough or invoice, depending on trade.
Timing also shapes the content of the review itself, not just whether it happens. A review written the same day tends to be specific: the crew's names, what got fixed, how clean the site was left. A review written three weeks later, if it happens at all, tends to be generic: "good job, would recommend." Specific reviews carry more weight with the next homeowner reading them, and they carry more weight with AI answer engines pulling review content, because specificity reads as authenticity.
For trades with a longer project arc (a full remodel, a multi-day roof replacement), the ask still belongs at the moment of completion, not partway through, and not after the invoice has been sitting unpaid for a month waiting on a change-order dispute. If there's an open issue at completion, the ask waits until that's resolved. Sending a review request into an active dispute is the fastest way to guarantee the response you didn't want.
- The highest-response window is same-day to 48 hours after completion
- Specific reviews (names, details) carry more weight than generic ones, with homeowners and with AI citation
- Multi-day jobs still get the ask at final walkthrough, not partway through
- Never send a review ask into an open dispute or unresolved complaint
Mistake #5: No system, so it depends on someone remembering
Most contractors we look at don't have a review process. They have a review habit that lives in one person's head, usually the owner's, and it runs whenever things are slow enough to remember it. That works for a few weeks after a New Year's resolution to "get more reviews this year" and then quietly stops the first busy month, which for most trades is most months.
A system means the ask is tied to a trigger that happens automatically: job marked complete in the scheduling software, invoice paid, final walkthrough signed off, whatever event already exists in the workflow. It goes out the same way every time, gets tracked (who was asked, who responded, who needs a second nudge), and someone owns checking the results weekly, not "whenever." It also means new reviews get monitored daily, not discovered by accident when a customer mentions seeing one.
The reason this matters more than it sounds: rating and review count are both moving targets, and a competitor running a consistent system adds reviews every week while a business running on memory adds them in bursts and droughts. Over a year, the compounding gap between five reviews a month and zero-then-fifteen-then-zero is the difference between a profile that looks active and current and one that looks stalled, even if total lifetime review count is similar.
- An ask that depends on someone remembering will lapse during every busy season
- Tie the ask to an existing workflow trigger: invoice paid, job closed, walkthrough signed
- Track who was asked and who responded; monitor new reviews daily, not by accident
- Steady weekly review flow reads as an active business; bursts and droughts read as stalled
Mistake #6: Panicking on a 1-star instead of handling it like a process
The instinct on a fresh 1-star is either to argue in the replies, to ignore it and hope it scrolls down, or to try to get it removed. Arguing in public makes the next reader side with the customer by default, even when the contractor is right. Ignoring it leaves an unanswered complaint sitting at the top of the profile (see Mistake #1). And trying to get it removed only works if it actually violates Google's policies (fake account, off-topic content, hate speech, conflict of interest), which most negative-but-real reviews do not.
What holds up: a same-day calm response acknowledging the specific complaint, a factual (not defensive) account of what happened if it's inaccurate, and a direct invitation to call and resolve it offline. If the complaint is legitimate, saying so and stating what was done about it does more for the next reader than any amount of arguing. If it's from a competitor, a disgruntled ex-employee, or someone who was never actually a customer, that's the one case worth flagging to Google for removal, since it may qualify as a policy violation rather than a genuine service complaint.
The other piece contractors underestimate: one or two 1-stars against a base of 150+ genuine reviews barely move the average and barely move a homeowner's decision. One or two 1-stars against a base of 20 total reviews move both a lot. The real fix for a single bad review's outsized weight usually isn't fighting that review. It's building enough volume of real reviews (see Mistake #2 and #5) that any single complaint gets diluted to where it belongs.
- Public arguing with a reviewer almost always reads badly to the next viewer
- Removal requests only work for genuine policy violations, not unhappy-but-real customers
- A calm, factual, same-day response does more work than any dispute
- Volume dilutes outlier reviews; a thin review base makes every 1-star hit harder
Mistake #7: Treating reviews as a rating, not a citation
For years, reviews were a trust signal for a human scrolling the map pack: a star count and a snippet or two. That's still true, but it's no longer the whole picture. When a homeowner asks an AI answer engine "who's a good septic company near me" or "best locksmith for a lockout," the engine is pulling from review text, review recency, and how a business responds to reviews, not just the star average, to decide what to cite and how to describe the business.
That changes what "good" review content looks like. A review that just says "great service, 5 stars" gives an AI engine almost nothing to work with. A review that says "they replaced our drain field in one day and left the yard cleaner than they found it" gives it a specific claim to cite. Contractors who coach customers toward specifics (not scripting them, just prompting with a question like "what job did we do for you?" in the ask) end up with a review base that reads well to both humans and answer engines.
The other half is response consistency. A profile where the owner replies to reviews with specific, on-brand language over time builds a pattern an AI engine can recognize as an active, accountable business. A profile with three years of silence in the owner-response column reads as dormant, regardless of the star average sitting on top of it. Reviews stopped being decoration a while back. They're now one of the source documents the internet's newest research assistant reads before it tells a homeowner who to call.
- AI answer engines cite review content and owner-response patterns, not just star averages
- Specific review content (job type, detail) gives an AI engine something concrete to cite
- Consistent, specific owner responses read as an active, accountable business over time
- A high star average with years of response silence still reads as a dormant profile