Why Review Recency Matters for Local SEO More Than Your Review Count Does

Most businesses treat reviews as a scoreboard. Get the number up, get the average above 4.5, move on. That thinking produces profiles with four hundred reviews where the newest one is fourteen months old.

A dormant profile is a strange thing to look at. The customer wonders whether the place has closed. A ranking system trying to work out which listings are still trading has to wonder the same thing. Count tells you about the past. Age tells you about now.

Why does review age carry information that review count cannot?

Count is cumulative and never decays. A business can gather four hundred reviews over a decade, stop entirely, and the count stays at four hundred forever. Age is the only part of a review set that changes without anyone doing anything, which makes it a usable proxy for whether a business is still operating.

Local search has a verification problem that organic web search does not. A blog post that was accurate in 2019 is usually still roughly accurate. A dry cleaner that was open in 2019 may be a vape shop now. The signals available to resolve that are thin: whether the website still responds, whether anyone has edited the listing, and whether real people are still turning up.

Of those, reviews are the one signal that comes from outside the business and cannot be fully controlled by it. A business can update its own hours on a Tuesday afternoon. It cannot make a stranger write about a visit that did not happen. So when practitioners talk about freshness in a local context, they mean whether the listing shows evidence of recent human contact.

What does a steady trickle of reviews signal that a burst of forty does not?

A handful of reviews arriving every month looks like ordinary trade. Forty arriving in one week looks like a campaign. The difference is not the total; it is the shape. Natural review velocity tends to track foot traffic, which is uneven but rarely spikes by an order of magnitude and then flatlines.

Consider the underlying reality. A dental practice sees a fairly consistent number of patients each week, and a small, stable proportion leave feedback. The pattern is lumpy, seasonal perhaps, but continuous. If that practice produces forty reviews in six days and nothing for five months, the honest explanations are a one-off push, an incentive scheme, or purchased reviews. None describes ongoing demand.

There is a second cost that gets overlooked. A burst ages as a block. Twelve months on, every one of those forty reviews is stale simultaneously. A business that collected three a month for the same period has a rolling window of recent customer reviews that never fully empties. The steady approach produces a better profile at every point in time, not just at the end.

When in the customer journey should you actually ask?

Ask at the point where the customer has received the thing they came for and is still in contact with you. For most service businesses that is the moment of handover or collection, or within a day of it. Asking a week later means competing with everything that happened in between.

The practical failure in most review programmes is not unwillingness to ask. It is asking at an administratively convenient moment rather than an emotionally relevant one. The invoice email goes out three days after the job because that is when accounts process it, so the review request rides along with it. By then the customer has moved on.

A few patterns that tend to hold across sectors with a physical presence:

• Ask the person who did the work. A request from the technician who fixed the boiler converts better than one from a no-reply address, because there is a face attached to it.
• Ask once in person, then follow up once in writing. The verbal ask sets the expectation; the written follow-up carries the link. Neither works as well alone.
• Remove every step you can. Link directly to the review form, not to the homepage of your Google Business Profile. Each additional click costs you people who intended to comply.
• Do not script the content. Suggesting what someone should write produces reviews that read identically, which is obvious to human readers and trivially detectable.
• Do not filter who you ask. Selecting only the customers you expect to be happy is treated as a policy violation on several platforms, and it produces a ratings profile that does not match reality anyway.

Does responding to reviews keep a profile active when new ones are slow?

Responses are owner-generated content attached to a listing, and they are timestamped. They do not substitute for new reviews, but they do mean a profile is never entirely without recent activity. For a business with genuinely low review volume, that matters more than it does for a busy one.

The case for responding is mostly not about ranking. A reply is read by everyone who looks at that review afterwards, which over the life of a listing is far more people than the one who wrote it. A measured reply to a two-star review about a long wait does more for a prospective customer’s confidence than the review does to damage it. An unanswered complaint reads as an uncontested one.

Three habits practitioners tend to converge on. Reply to negative reviews within a day or two, briefly, without arguing the facts in public, and with a route to take it offline. Reply to positive ones too, but keep it short and avoid the copy-paste block, because fifteen identical thank-you messages in a row look automated. And never reply when annoyed, because that reply is permanent and screenshots travel.

For a solicitor’s office or a specialist clinic collecting one review a month, a consistent response habit is the difference between a listing that shows activity in the last fortnight and one that shows activity from the spring.

What happens when one branch of a multi-location business stops collecting?

It falls behind the others, quietly, and usually nobody notices for two or three quarters. Each location has its own listing and its own review set, so recency does not pool across the group. A strong brand average provides no cover for the branch whose newest review is from last year.

This is the common version of the problem in businesses large enough to have a marketing function. Head office reports in aggregate: total count across the estate, group average rating, both trending up. Underneath, two branches are doing almost all the collecting and four have gone quiet, usually because the manager who cared about it moved on.

The fix is reporting, not strategy. Track the date of the most recent review per location, not just the count and average, and put it in front of whoever runs each site. The moment a branch can see that its newest review is eleven weeks old while the one down the road is at four days, the behaviour tends to correct itself. It is the aggregation that hides the problem.

How much does recency actually count for?

Nobody outside Google knows. Recency appears to function as one input among many, alongside proximity, relevance, the quality and consistency of listing data, links, and the rest. Anyone quoting you a percentage weighting for it is guessing, and the weighting almost certainly varies by query type and sector.

What practitioners can say is narrower. Profiles that stay active tend to hold their local search visibility better than comparable profiles that go dormant. Businesses that resume collecting after a long gap usually see things improve, though rarely immediately and rarely in a way you can separate from everything else they changed. And in competitive categories, the businesses with current reviews are more often the ones ranking.

That is correlational, observed across accounts, and subject to the usual problem that nobody runs controlled experiments on their own revenue. The honest framing is that review recency is cheap, under your control, improves conversion independently of ranking, and carries almost no downside risk. You do not need its exact weight to justify doing it.

Why does buying reviews fail on both the ranking side and the trust side?

Purchased reviews fail twice. The platform side is a detection problem the buyer cannot win, because the patterns are statistical and the seller is optimising for volume. The customer side is worse: fake reviews read like fake reviews, and a reader who suspects one stops trusting the genuine ones too.

Take detection first. Bought reviews arrive in clusters, from accounts with thin histories, often from the wrong geography and in a vocabulary that does not match how real customers describe that kind of business. Google removes large volumes of policy-violating reviews and has added penalties for profiles involved in it. Enforcement is imperfect, which is why it is a bad bet: it may miss you this quarter and remove half your review set the next.

The trust side gets underweighted. People have become reasonably good at spotting manufactured praise. Five reviews in a row that name the business in full, praise “the team” in general terms and contain no specific detail about what actually happened will register as wrong even to someone who could not explain why. Once that suspicion forms, it contaminates everything on the listing, including the real reviews you worked for.

Key takeaways

• Count never decays; age is the only part of a review profile that changes on its own, which is what makes it a freshness signal.
• A steady trickle tracks real trade. A burst of forty in a week looks manufactured and ages as a single block.
• Ask at handover or within a day of it, once verbally and once in writing, and link straight to the review form.
• Replying keeps a listing showing recent activity when new reviews are slow, and is read by far more people than the reviewer.
• In multi-location businesses, track the newest review date per branch; group averages hide the branch that stopped collecting.

Frequently asked questions

How old does a review have to be before it counts as stale?

There is no published threshold, and anyone offering one is inventing it. As a working rule, practitioners get uncomfortable when the newest review passes about three months, because at that point a customer reading the profile starts to wonder. Judge it against your competitors in the same category rather than against an absolute number.

Can you ask customers to update an old review instead of writing a new one?

You can, and some platforms timestamp the edit, which refreshes it. In practice the response rate is poor, because you are asking someone to find an account they may not have used in a year. It is a reasonable thing to try with regular clients, but it is not a substitute for a collection habit that generates new reviews.

Do reviews on other platforms help local search visibility?

They are widely believed to contribute, and they certainly contribute to how a business looks when someone researches it. The weighting given to third-party platforms is unknown and probably varies by sector, with industry-specific sites mattering more in regulated fields. Most businesses are better off keeping one or two secondary platforms current than spreading thinly across six.

Is it worth responding to reviews that are several years old?

Usually not individually, with one exception. If an old negative review is still visible near the top of your profile and has never been answered, replying now is worth doing, because new readers see it unanswered. Beyond that, a sweep of replies to ancient reviews is an unusual activity pattern and better effort goes into current ones.

None of this is difficult. It is just unglamorous, and it sits with operations rather than marketing, which is why it slips. The businesses that do it well have built one small habit into a routine that already exists: the engineer asks at handover, the receptionist checks the listing on Mondays, somebody reports the newest review date per branch. No campaign, no budget line. Just a profile that keeps showing the business is still there.

About the author

This article was written by the digital marketing team at TheGenieLab, a UK-based agency working with growing retail and service brands since 2011. They advise on search, conversion and the operational side of keeping a business findable. More about their work is at TheGenieLab.

FAQPage schema

{
  “@context”: “https://schema.org”,
  “@type”: “FAQPage”,
  “mainEntity”: [
    {
      “@type”: “Question”,
      “name”: “How old does a review have to be before it counts as stale?”,
      “acceptedAnswer”: {
        “@type”: “Answer”,
        “text”: “There is no published threshold, and anyone offering one is inventing it. As a working rule, practitioners get uncomfortable when the newest review passes about three months, because at that point a customer reading the profile starts to wonder. Judge it against your competitors in the same category rather than against an absolute number.”
      }
    },
    {
      “@type”: “Question”,
      “name”: “Can you ask customers to update an old review instead of writing a new one?”,
      “acceptedAnswer”: {
        “@type”: “Answer”,
        “text”: “You can, and some platforms timestamp the edit, which refreshes it. In practice the response rate is poor, because you are asking someone to find an account they may not have used in a year. It is a reasonable thing to try with regular clients, but it is not a substitute for a collection habit that generates new reviews.”
      }
    },
    {
      “@type”: “Question”,
      “name”: “Do reviews on other platforms help local search visibility?”,
      “acceptedAnswer”: {
        “@type”: “Answer”,
        “text”: “They are widely believed to contribute, and they certainly contribute to how a business looks when someone researches it. The weighting given to third-party platforms is unknown and probably varies by sector, with industry-specific sites mattering more in regulated fields. Most businesses are better off keeping one or two secondary platforms current than spreading thinly across six.”
      }
    },
    {
      “@type”: “Question”,
      “name”: “Is it worth responding to reviews that are several years old?”,
      “acceptedAnswer”: {
        “@type”: “Answer”,
        “text”: “Usually not individually, with one exception. If an old negative review is still visible near the top of your profile and has never been answered, replying now is worth doing, because new readers see it unanswered. Beyond that, a sweep of replies to ancient reviews is an unusual activity pattern and better effort goes into current ones. None of this is difficult. It is just unglamorous, and it sits with operations rather than marketing, which is why it slips. The businesses that do it well have built one small habit into a routine that already exists: the engineer asks at handover, the receptionist checks the listing on Mondays, somebody reports the newest review date per branch. No campaign, no budget line. Just a profile that keeps showing the business is still there.”
      }
    }
  ]
}