By Rahul Saini, Author at Search Counsel Co. Last updated [JULY] 2026.

Most businesses treat reviews as a scoreboard: get the number up, get the stars up, done. That’s why so many of them run a review push, hit 60 reviews, stop, and watch nothing happen. The local algorithm isn’t reading your total. It’s reading your pattern, and it’s reading what you say back.

1) What review velocity is

Review velocity is the rate at which new reviews arrive, measured over time rather than in total. Three reviews a week, every week, is a velocity. Sixty reviews in March and nothing since is not a velocity, it’s an event.

Google has never published a velocity metric, and anyone who tells you the exact threshold is guessing. What we can say from the ranking data and from how enforcement behaves is that the algorithm reads pattern as much as volume: how recent your reviews are, how consistently they arrive, and whether the pattern looks like a business serving customers or a business running a campaign.

2) Why recency overtook volume

This is the single most useful thing in the current data. In Whitespark’s 2026 Local Search Ranking Factors survey, review recency climbed from around 93rd place in 2023 to around 11th, the largest positional jump of any factor in the study. Review signals overall now carry roughly 20% of local pack weight, up from about 16% in 2023, and that’s the biggest gain of any signal group.

The logic is obvious once you say it out loud. A restaurant with 400 reviews, none since 2023, tells Google nothing about whether it’s any good now, or whether it’s even open. A restaurant with 90 reviews and eight from last month is demonstrably alive and demonstrably still serving people.

3) The burst problem

The instinct, once you learn recency matters, is to run a campaign. Get forty reviews this month, fix the recency problem, move on.

Don’t. A spike is a spam pattern. Google’s systems look for exactly this shape, and a sudden cluster of reviews on a profile that had been quiet is one of the clearest signals of a review push, whether or not the reviews are real. Profiles caught in a fake-engagement sweep can have reviews unpublished, lose the ability to receive new ones for a period, and in some cases display a public warning telling consumers that fake reviews were removed.

A slow drip is safer, and it’s also just better. It produces a review profile that keeps working instead of one you have to keep resuscitating.

4) What pace should you aim for?

There’s no universal number, and any article giving you one is inventing it. The benchmark is competitive.

1. Open your map pack. Look at the three businesses currently occupying it for your main term.
2. Note their total review count and, more to the point, the date on their most recent few reviews.
3. Work out their approximate pace. Two a month? Ten a week? That’s your target range.
4. Set a pace you can sustain forever, not one you can sustain for a quarter. Consistency is the entire point. Eight a month for three years beats a hundred in one summer.
5. Tie it to a natural trigger. Every completed job, every appointment, every invoice. If the request is a step in your process, the velocity takes care of itself. If it’s a marketing task someone has to remember, it dies in week three.

For most small local businesses, a request sent to every customer after every transaction, with a realistic 5% to 15% conversion rate, produces a steady flow without any campaign at all. That’s the whole system.

5) Responses: the lever nobody pulls

Most businesses never reply to reviews. That’s a strange thing to leave on the table, because responding does three separate jobs at once.

• It’s an engagement signal. A profile where the owner consistently responds looks like an actively managed, legitimate business. That’s exactly the signal Google’s spam-conscious systems are looking for now.
• It’s indexed text you control. This is the finding most people miss. Sterling Sky’s controlled testing found that keywords in the review text customers write have no impact on rankings, but the text of your response is indexed and does contribute as a relevance signal. Your reply is the only part of a review you get to write, and it’s the part that counts.
• It converts the reader. Prospective customers read your responses. A thoughtful reply to a bad review persuades more people than the review discourages.

6) How to write a response that works

The trick is to be specific without being robotic, and to let the specificity do the SEO work naturally.

Notice what’s happening in the right-hand column. The service is named. The area is named. The response reads like a person wrote it. None of that is keyword stuffing, and all of it is relevance.

7) Responding to negative reviews

A bad review is not a crisis. A bad review with no reply, sitting at the top of your profile for eight months, is.

The structure that works:

1. Acknowledge the specific thing. Not “we’re sorry you feel that way,” which reads as contempt. Name what went wrong.
2. Take the part that’s yours. Even if they’re mostly wrong, there’s usually something.
3. Say what’s changed. This is the sentence future customers actually care about.
4. Offer a route back, once. Then stop. A public back-and-forth helps nobody.

Write it for the next hundred people who’ll read it, not for the person who wrote it. They’ve likely gone. The audience hasn’t.

If a review is truly fake, off-topic, or from someone who was never a customer, report it. That process is slow and inconsistent, and it’s still the right move. What you can’t do is offer an incentive to have it revised or removed.

8) Does a perfect 5.0 help or hurt?

A flawless rating with no dissent is a weaker signal than most people assume, and consumers have learned to distrust it. It reads as filtered, because historically it usually was.

The useful goal isn’t perfection. It’s a strong average with enough volume and enough recency to be credible, plus visible evidence that you handle the occasional complaint like an adult. A well-answered three-star review does more for a prospective customer’s confidence than the fortieth identical five-star.

9) Reviews beyond Google

Google reviews are the ones that move the map pack. They are not the only ones that matter.

The 2026 survey’s AI-visibility section scored the diversity of third-party sites carrying your reviews as one of the stronger signals. A business with reviews only on Google is a single point of evidence. A business reviewed on Google, on the industry platform its customers actually use, and on a general review site is corroborated from three directions, which is precisely what an AI assistant needs before it will recommend anyone.

Find the two or three platforms your industry truly uses, claim your profile, and let the flow build there too. Related: the best local citation sources and how to build them.

10) How reviews feed AI recommendations

Google reads review content, not just star ratings. So do the assistants people now ask for local recommendations, and Google is replacing the old profile Q and A section with AI-generated answers assembled from reviews, posts, and your web presence.

This changes what a good review looks like. A profile of coached, identical, six-word five-star reviews gives a model nothing to work with. A profile where customers describe the specific problem you solved, in their own words, hands a model the exact language it needs to recommend you for that problem.

Which means the unglamorous, compliant approach isn’t a compromise. Unscripted reviews describing real jobs are simultaneously the safest thing to collect, the most persuasive thing for a human to read, and the most useful thing for a machine to cite. Check where you currently stand with our free Local AI Visibility Grader.

11) What to measure

Stop reporting total review count. Report these instead:

• New reviews per month, plotted as a line. You want a flat, unremarkable line that never touches zero.
• Days since your last review. If this number ever passes 30, your process has broken.
• Response rate and response time. Target 100% and under 48 hours.
• Your pace against the top three in your map pack. The only benchmark that means anything.
• Review count on non-Google platforms. The diversity signal.

A response library you can actually use

Your reply is the only indexed text in the exchange that you control. Vary the wording, name the service and the area naturally, and never sound like a template, which means never using a template verbatim. Use these as shapes, not scripts.

12) Sources used for this guide

FAQ: review velocity and responses

Conclusion: pace beats pile

The review count on your profile is a historical record. Your velocity is a live signal, and your responses are the only part of the whole system you get to write. Between them they carry a fifth of what decides the map pack, and they’re entirely within your control.

Set a pace you can hold forever, wire the request into a step you already perform, and reply to everything within two days. That’s it. That’s the discipline. If your collection process still involves a tablet at reception or a script asking customers to name someone, fix that first: how to get more Google reviews without breaking Google’s rules.