Google review velocity vs review count: what local SEO actually rewards in 2026

One disclaimer before we go any further. Nobody outside Google's ranking team knows exactly what Google weights, in what proportion, for local search. What follows is pattern inference from watching thousands of business profiles over the last few years. It's evidence-based, but it isn't the official answer because there is no official answer. Trust the directional shape, not the exact percentages.

Why count alone stopped working

Imagine two competing dental practices. One has 380 reviews, the last one from 14 months ago. The other has 110 reviews, the last one from 3 days ago. Until roughly 2020, the practice with 380 reviews dominated local search. Today, in most metro markets, the practice with 110 reviews wins the local pack. The reason is recency.

Recency works as a signal because Google's job is to rank the businesses that are currently good, not the ones that were good five years ago. A profile with 380 stale reviews looks like a business that stopped trying. A profile with 110 fresh reviews looks active and trusted right now. The 110-review business gets the calls.

The four signals that actually matter

Across the businesses we monitor, four signals correlate consistently with local pack ranking movement. Each one is independently controllable — that's what makes the framework useful.

Velocity — the most important signal

Velocity is the rate at which new reviews land. It's the single signal we see the strongest correlation with ranking changes. A practice that goes from 1 review per month to 6 per month tends to move up the local pack within 60-90 days, even before total review count changes meaningfully.

The target velocity depends on your visit volume. A dental practice seeing 200 patients a month should aim for 5-10 reviews a month. A solo plumber doing 80 jobs should aim for 3-5. The right rule of thumb: you should be converting somewhere between 2 and 5 percent of customer interactions into reviews. Below 2 percent and you're leaving signal on the table. Above 5 percent and you risk the kind of velocity spike that looks unnatural (more on that below).

The other reason velocity matters: it lets Google distinguish "currently active business" from "historically popular business." That distinction is increasingly weighted as the algorithm tries to surface businesses that are operating well right now.

Recency — the silent killer

Recency is the signal that most teams don't notice has decayed. A business with 200 reviews and a 90-day gap since the last one looks worse to Google than a business with 60 reviews where the most recent landed yesterday.

The threshold varies by category. Active service businesses (restaurants, plumbers, dentists, stylists) should have a new review every 7-14 days. Less-frequent service businesses (contractors, specialty doctors, financial advisors) can tolerate 21-30 days. Beyond those windows, Google's algorithm starts to treat your profile as stale.

The good news: recency is the easiest of the four signals to recover. One consistent same-day review-request workflow can move a profile from "stale" to "fresh" in two weeks. We've watched businesses move multiple ranks in the local pack purely from fixing recency.

Response rate — the trust signal

Google has said publicly that responding to reviews increases trust signals on your profile, and our observations back that up. But the dynamic isn't just about quantity of responses — it's about which reviews you respond to and how fast.

Negative reviews matter more than positive ones for response signaling. A profile that responds to 100% of its 1- and 2-star reviews within 48 hours signals an attentive business. A profile that ignores its negative reviews (even if it responds to every positive one) signals disengagement.

On positive reviews, you don't need 100% coverage. 70% is fine, especially if the responses are specific rather than canned. Generic "Thanks for the kind words!" responses actually appear to do less than no response at all — they signal automation.

Variance — the authenticity signal

This is the signal teams optimize the least but trip on the most. Variance refers to the diversity of your review program — different reviewers, different review lengths, different phrasing, different times of day, different platforms across the customer journey.

A pattern of 47 reviews that are all 2-3 sentences long, all left between 11am and 1pm, all from reviewers with two or fewer prior reviews — that's the signature of a review-request workflow that's too uniform. Google's spam systems are increasingly good at flagging this kind of pattern. Even if the reviews are legitimate (and most of them are), the pattern itself is the problem.

The fix is to introduce variance deliberately. Stagger when you send review requests across the day. Use different scripts for different moments of the customer journey. Don't put a QR code at one specific touchpoint that funnels every review through the same channel. Variance is the unsexy maintenance work that protects you from algorithm changes.

What this means for the way you run your program

If you take three things from this article, take these.

One. Stop reporting on total review count. Start reporting on monthly velocity and recency. Track both as monthly metrics with a 90-day rolling baseline. These are the numbers your local SEO actually depends on.

Two. Build a same-day review-request workflow. The recency signal is the cheapest one to fix and the most impactful. Most businesses we work with see ranking lift within 60 days of putting same-day requests in place.

Three. Respond to every negative review within 48 hours. The response is for future customers more than for the algorithm, but the algorithm rewards it too. Don't use generic responses — those count against you, not for you.

What about overall star rating?

Stars still matter. A 4.8 outperforms a 4.3 in most categories. But the marginal value of going from 4.7 to 4.9 is much smaller than the marginal value of going from 1 review per month to 6 per month. If you have a 4.5+ rating, your ranking constraint is almost certainly velocity and recency, not the rating itself.

One nuance worth knowing: extremely high ratings (4.95+) with very high volume can actually trigger Google's "too good to be true" review filtering, especially in categories where 4.5-4.7 is more typical. If you're running a 4.97 with 800 reviews, expect some of them to be quietly suppressed.

The honest summary

Total review count is a vanity number. Velocity, recency, response rate, and variance are the operational numbers. Build the program around those, measure them every month, fix recency first, build a response habit second, vary your sources third. Do that for six months and your local pack ranking will reflect it. That's the whole architecture.