A hundred five-star reviews from three years ago will lose to thirty from the last few months, and it isn’t close. Most contractors built their review count during a specific push — a hiring drive, a new office manager who cared about it, a slow month where someone finally sat down and asked — and then let it go quiet. The count looks great. The trend line doesn’t, and increasingly, the trend line is what both customers and Google are actually weighing.
Start with what customers say they do, because it’s more specific than most owners assume. In BrightLocal’s Local Consumer Review Survey, a large and growing share of consumers say they only trust reviews written within the last few weeks to a few months — not “read reviews,” specifically weight how recent they are before deciding whether to trust them at all. A stack of old reviews doesn’t get discounted a little. For a meaningful slice of people looking at your listing right now, it gets treated as if it barely counts, because it’s answering a question — “is this business good?” — with information that’s stale by the time they’re asking it.
Google’s own guidance points the same direction without spelling out an exact formula. Google’s stated local ranking factors are relevance, distance, and prominence, and prominence explicitly includes review signals — count and rating, but also how you handle them. Google doesn’t publish an exact recency weighting, and anyone who tells you a precise percentage is guessing. What’s not in dispute is the direction: a profile with a steady trickle of recent reviews reads as an active, currently-operating business in a way a profile that hasn’t moved in two years doesn’t, to a customer scanning it and, by extension, to how Google treats it too.
So the practical target isn’t “get to 100 reviews” — it’s “never let ninety days go by without new ones.” A business with thirty reviews, five of them from the last month, is showing a healthier signal than a business with two hundred reviews and nothing since a marketing push two years ago. That reframes review generation from a project you finish to a habit you run continuously, which is a harder sell internally but the honest one. It also means the job of “getting reviews” can’t sit with whoever happens to remember to ask that week — it needs an owner, someone whose job includes checking the last-thirty-days count the way they’d check any other number that slips if nobody’s watching it.
Asking without being obnoxious comes down to timing and specificity, not the wording of the request. Ask right after the job, while the technician is still there or within the hour — not three days later in a batch email that reads like it went to everyone. Ask the person who was actually home for the work, not whoever’s name is on the invoice if that’s a different person. And make it about the specific job, not a generic “how did we do” — a text that says “thanks for having us out for the AC repair today, mind leaving a quick review?” gets a different response than a form email from an address nobody recognizes. The businesses that get accused of being pushy about reviews are usually the ones asking everyone the same way at the same interval regardless of whether the job actually went well — which brings up the harder question of who not to ask.
Don’t ask a customer whose job didn’t go cleanly — a callback, a part that didn’t show, a price that came in higher than expected. That sounds obvious and gets skipped constantly, because the review-generation tool or the checklist doesn’t know the difference between a smooth job and a rough one; it just knows a job closed. A blanket ask-everyone system will occasionally ask someone who’s annoyed at you to leave a review, and you will not like what they write.
Which gets to the bad review itself, because one is coming eventually no matter how good the work is. Respond to it, briefly, without getting defensive in public — acknowledge what happened, state plainly what you did or would do about it, and take the detailed back-and-forth offline. Future customers reading it are watching how you handled it more than they’re weighing the original complaint; a calm, specific response to a bad review is often more persuasive to a stranger than another five-star review would have been, because it shows how you behave when something goes wrong instead of only when everything went right. What doesn’t work is arguing with the reviewer in the replies, or ignoring it and hoping it scrolls down — both read as confirmation to anyone comparing you against a competitor’s profile.
One honest exception: if you’re already getting a steady stream of recent reviews and responding to what comes in, more volume past that point has sharply diminishing returns. Going from thirty recent reviews to three hundred doesn’t move a homeowner’s decision nearly as much as going from zero recent reviews to a few active ones does. At that point, the review side of your marketing is not where your growth is sitting, and it’s worth looking at whether the leak is somewhere earlier — the funnel piece that opens this series covers the more common places that turns out to be true.
Reviews are also one of the few marketing assets that keep doing something for you in the months your phone would otherwise go quiet — which is worth understanding on its own terms, in the next piece, on what actually fills the slow months.