Industries

Do reviews help local SEO? What actually moves the map pack

20 October 2026 · JustReview

Do reviews help SEO? For local search, yes: Google says in its own help pages that the number of reviews a business has and its rating feed into how it ranks on Google Maps and in the local pack. They are one input to one of three factors, though, and they do nothing on their own for a business listed in the wrong category or far from the person searching. Below is what Google confirms, what it does not, and what our data on 779 local Google profiles and a sample of 1,756 written reviews, measured on 20 October 2026, says about the parts a business can still change.

What Google actually confirms

Google’s help page on local ranking, which we read on 20 October 2026, says local results are “mainly based on relevance, distance, and popularity”, and then describes the third factor under the heading of prominence. Of that factor it says it is “based on info like how many websites link to your business and how many reviews you have”, and that “more reviews and positive ratings can help your business’s local ranking”. That is the whole official statement on reviews and ranking. The same page tells owners to respond to reviews, but frames it as standing out, not ranking higher.

Most of what is written about reviews and local SEO goes further than that. The table sets the common claims against what Google itself says, ordered from the most supported to the least.

Industry surveys of local SEO practitioners put review recency and review wording high on their lists, and they may be right. But a survey measures what experts believe, and Google does not publish its weights. What we can add is the other half: what the review signals of ordinary local businesses look like in practice.

Count and rating: real factors that move slowly

The 779 profiles in our sample belong to businesses that connect exactly one Google Business Profile to JustReview and no shop or marketplace source, with a working import and at least one review in the past two years, so online shops and multi-location chains are left out. The middle profile among them shows 44 reviews in total and collected 6 in the past 12 months. One in four collected 17 or more, and 87 of them (11.2%) collected nothing at all.

The count Google shows is therefore mostly history. For the median profile, only 15% of the visible total was written in the past year. A competitor with 300 reviews will not be caught by a good month, and a business that stops asking does not see its number fall, so it rarely notices the slowdown. We looked at the pace side of this in detail in how fast normal businesses collect reviews, and at what the total needs to be in how many reviews a business needs.

The rating is a weaker lever than people expect, for a simple reason: most local profiles are already near the top. When we measured displayed ratings, the median single-location profile with 25 or more reviews showed 4.9 and almost nine in ten showed 4.5 or higher. When every competitor in the pack shows a rating in the high fours, the rating stops separating them and the count, the distance and the category do the work.

Freshness: the gap most local profiles have

Google does not name review recency as a ranking factor. A customer sees it, though, because every review carries a date, and “a year ago” on the newest review reads as a business that has gone quiet. On that measure the 779 profiles look worse than their totals suggest.

On the morning we measured, the middle profile had last been reviewed 50 days earlier. Even the businesses that collect steadily are not immune: of the 274 profiles that gathered at least 12 reviews in the past year, 58 (21.2%) had gone more than 30 days without one. Reviews arrive in bursts, after a busy week or a push to ask, and then stop.

If recency matters for ranking, this is where most local businesses lose ground without knowing it. If it does not, it still matters for the click, because the map pack shows the rating and count, and the profile behind it shows the dates.

Keywords in reviews: what the reviews actually say

The most repeated piece of local SEO advice about reviews is to get customers to mention your service and your town, on the theory that Google reads review text for relevance. Google has never said this. Google does pull short phrases out of reviews and show them under some local results, which is why the advice feels plausible, but showing a phrase is not the same as ranking on it.

We measured the part that can be measured: how often real reviews contain the words the theory needs. First, the reviews that contain no words at all. Of the 21,842 Google reviews our local profiles received in the past 12 months, 4,048 (18.5%) were a star rating with no text. Those reviews count toward the total and the average, and they cannot carry a keyword.

Then we drew a random sample of written reviews, up to three per business, giving 1,756 reviews from 672 businesses in more than 20 languages, and had each one labelled for four things.

Service words do turn up: roughly six in ten written reviews mention what the customer came for. That is the loosest of the four labels, and our spot check found it generous on short reviews, so read it as an upper bound. The words the keyword theory leans on most are rarer. 29.2% name the kind of business, 8.4% name a town or district, and only 4.4% contain both, the “best dentist in Leeds” pairing that review templates try to produce. Count the reviews with no text as well and about one review in four names the business type at all, while fewer than one in twenty pairs it with a place. What reviews mention far more often is people: 43.0% of five-star reviews name a member of staff, against 13.2% of the 91 one- and two-star reviews in the sample.

The practical conclusion is not that wording is worthless. It is that you cannot engineer it. Reviews that read like a template are what review filters are built to catch, and copied phrasing is the signature we found when we looked at what purchased reviews have in common. What does work is asking a better question. “How was your appointment?” gets “Great!”; “What did we do for you today?” gets a sentence that names the treatment, the repair or the dish, in the customer’s own words.

Replies: for the customer, not the algorithm

Google’s ranking page puts replies under a heading of their own and says that “positive reviews and helpful replies can help your business stand out”. It stops short of saying they lift your position. They are still worth the time. Someone choosing between three dentists in the pack opens the profiles, and the one that answers its one-star reviews calmly and specifically reads as the safer choice. Across the Google reviews we tracked between April and October 2026, owners answered 44.6%; the breakdown and what good answers contain are in our Google Business Profile reviews guide.

Do reviews on your own website help SEO?

This is a separate question, and the honest answer is: not for the map pack. Google ranks your Business Profile from what it knows about the profile, and showing those same reviews on your website does not add to the count or the rating there. On the organic side, Google does not show review stars for a business reviewing itself on its own site, which we cover in our guide to review schema, and many review widgets render in the browser, so the review text is not in the HTML a crawler first receives.

What reviews on your site do is finish the job the map pack started. The visitor who clicked through from Maps has already seen your stars; the page they land on should show the reviews that answer their next question, without sending them back to Google to read them.

What to do this month

1. Fix relevance first. Category, services and opening hours decide which searches you can appear for. Reviews cannot make up for the wrong primary category.
2. Ask every customer, at the same point every time. A steady trickle beats a campaign. Our guide to getting more Google reviews covers the direct link, the QR code and what Google does not let you offer in return.
3. Ask a question that invites detail. “What did we help you with?” produces service words on its own; a script does not.
4. Check the date of your newest review, not the total. If it is older than a month, the asking has stopped somewhere.
5. Answer the critical reviews first, then the rest, in your own words. Customers read replies to one-star reviews more than any other part of a profile.

Where we fit

JustReview helps with the parts of this that happen off Google. Our free Google review link generator gives you a direct link and QR code to the review form, so asking takes one tap. Connect your Business Profile and you can read and answer your Google reviews from the JustReview dashboard, and show them on your own pages with a badge, carousel or full reviews page that keeps up as new reviews arrive. The setup is on our Google reviews widget page, plans are on the pricing page, and you can create an account and check the date of your newest review in a couple of minutes.

To be clear about the limits: nothing in a widget changes your position in Google Maps. Ranking comes from your profile, your category, your distance from the searcher and the reviews Google holds there. What we do is make the review collection habit easier to keep and make the reviews you already have work on your own site.

How we measured

All figures come from JustReview’s production database on 20 October 2026, read-only. The profile sample is every account with exactly one Google Business Profile, connected through Google’s API with a valid token, and no shop or marketplace source, excluding our own test accounts and accounts where a plan limit pauses the import. Recency counts the 779 of those with at least one review dated in the past two years. The text sample took up to three random written reviews per business from the past 12 months; each review was labelled by a language model against fixed definitions, and a person read 45 random labels against the review text and disagreed with four, each time a label that said yes where no fitted better. Business names are not published.