How to Embed Google Reviews on Your Website: 3 Routes, 7 Steps, and What Each One Costs

Most businesses install a Google reviews embed for the wrong reason. They have been told the stars will show up in search results, so they add a widget, wait a month, and see nothing change. Google's own documentation is unambiguous about why: if the entity being reviewed controls the reviews about itself, its pages carrying LocalBusiness or Organization markup are ineligible for the star review feature, and Google names embedded third-party widgets such as Google Business reviews as the exact case it means.

That does not make the embed worthless. It relocates the value. Reviews on your own site are a conversion asset, not a search asset. Somebody who arrives from an ad, a van, a business card or a referral never sees your Google listing, and a decade of earned goodwill is invisible to them. Closing that gap is worth doing properly, which means picking the route that matches your budget and your tolerance for maintenance rather than installing the first widget that appears in a search.

There are three routes. Seven steps below take you from a claimed listing to a working, measured embed. Do them in order.

Every WebForger site ships with the Google Reviews app switched on at no extra cost, along with every other app we build. If you would rather not run this project yourself, tell us about your site and we will show you what it looks like on yours first.

Step 1: Confirm the listing is claimed and the reviews are worth showing

Nothing downstream works if the source is not in order. Sign in to Google Business Profile and confirm the listing is verified and that you are an owner or manager, not just a contributor. An unclaimed listing still collects reviews and still renders publicly, but you cannot reply to reviewers, you cannot flag anything, and several of the tools in Step 4 will not connect to it.

Then look honestly at what you have. A 4.3 average from 140 reviews is a strong asset and worth putting on the homepage. A 4.9 average from six reviews is not persuasive to anyone, because visitors read low volume as "friends and family" whether or not that is fair. Under roughly twenty reviews, the higher-value project is collecting more of them, and the embed can wait a quarter.

Copy your Place ID while you are here. Google publishes a Place ID Finder, and almost every method below asks for that string rather than your business name. Save it somewhere you will find it again.

• Verify ownership of the listing before you evaluate any tool, because half of them fail at the connection step otherwise.
• Fewer than about twenty reviews means the embed will undersell you; fix the supply first.
• Record your Place ID once and reuse it; it is the input every route wants.

Step 2: Choose the route before you choose the tool

The three routes differ on one axis that matters more than features: who is responsible for keeping the reviews current. Manual screenshots put that on you forever. A third-party widget puts it on a vendor you pay monthly. A platform-native renderer puts it on the platform that already hosts your site.

Route one, pasting screenshots, suits a business that needs something on the page this afternoon and genuinely does not care whether it is still accurate in March. Route two suits a site on a platform with no first-party answer, where you can install arbitrary scripts and a small monthly fee is easier to approve than a build. Route three suits anybody whose site already runs on a platform that renders reviews itself, and it is the only one of the three where the visual result is under your design system's control rather than a vendor's.

Pick one now. Trying two in parallel is how sites end up with a stale screenshot halfway down the homepage and a live widget in the footer, disagreeing about the star rating.

• The real question is who maintains freshness: you, a vendor, or your platform.
• Route choice is a maintenance decision dressed up as a design decision.
• Commit to one route across the whole site so two versions of your rating never appear on one page.

Step 3: Route one, paste screenshots and accept the decay

Open your listing on a desktop browser, screenshot two or three individual reviews and the summary block with the star average, crop each one tightly, and place them on the page as images with descriptive alt text. Total cost is zero and total time is about twenty minutes. That is the entire appeal, and it is a real appeal for a one-page site that a builder wants to hand over on Friday.

The costs arrive later. The data is frozen the moment you save the file, so a business that earns two reviews a week is showing an increasingly wrong rating within a quarter, and visitors who check the listing find a different number. Screenshots are also images, which means the text inside them is not selectable, not translatable, not readable by a screen reader beyond whatever you put in the alt attribute, and not searchable. On mobile, a 1200 pixel wide screenshot of a desktop review card either shrinks until the text is unreadable or forces a horizontal scroll.

There is a credibility cost too, and it is the one people underestimate. A screenshot of a review is exactly as easy to fabricate as a paragraph of your own marketing copy, and visitors know it. If you take this route, at least set a quarterly reminder to redo the images, and link the block to your live Google listing so a sceptical visitor can verify the number themselves.

• Budget zero dollars and twenty minutes; this is the only route with no ongoing anything.
• Set a recurring quarterly reminder to retake the images, and expect to ignore it.
• Always link the block to the live listing, because a static screenshot alone proves nothing.

Step 4: Route two, rent a widget from a reviews vendor

This is the default answer, and it works. Elfsight, Trustindex, EmbedSocial and a dozen others all do the same thing: you connect your Place ID in their dashboard, choose a layout, and they hand you a snippet to paste into your page. The reviews refresh on their schedule, the layouts are competent, and installation takes about ten minutes on any platform that lets you paste raw HTML.

Price it before you install it. Elfsight's Google Reviews app, checked on 20 September 2026, offers a free tier capped at 200 widget views a month with Elfsight branding shown on the widget, then Basic at USD 5 a month for 5,000 views, Pro at USD 10 for 50,000, and Premium at USD 20 for 150,000, with enterprise tiers from USD 30 to USD 80. The free tier's 200 views is the number to look at closely: a site getting 1,000 sessions a month blows through it in the first week, after which the widget stops rendering. Competitors price similarly. Call it USD 60 to USD 240 a year, indefinitely, for a block of text on your own homepage.

The technical cost is the one nobody quotes you. The snippet loads a script from the vendor's domain, which means an extra DNS lookup, an extra TLS handshake and an extra render-blocking dependency on the page where people decide whether to call you. If the vendor has an outage, the block is empty. If the vendor changes its pricing or gets acquired, your homepage is downstream of that decision. Several of these widgets also render inside an iframe, which is what produces the familiar effect of a panel that clearly belongs to a different website, bolted onto yours.

Route two is the right answer when your platform gives you nothing better and you can paste a script. Go in with the yearly figure and the third-party dependency written down, not just the monthly one.

• Expect USD 60 to USD 240 a year for small business traffic volumes, forever, with prices confirmed on the vendor's own pricing page.
• Free tiers are view-capped and branded; 200 views a month is a fortnight for most real sites.
• Measure the added script weight in Step 7 before you accept the install.

Step 5: Route three, render the reviews from the platform your site already runs on

The third route removes the vendor from the diagram. Your site platform pulls the reviews from your listing on a schedule, caches them on its own network, and renders them into your page from its own domain. There is no second company in the chain, no monthly fee attached to a single block, and no script from an unrelated origin on your highest-intent page.

On WebForger this is the Google Reviews app, and it is included with every site rather than sold per month, as are the rest of the apps we build. The mechanics are worth stating precisely, because the honest version is more useful than the marketing version. The platform refreshes the review payload from your listing on a daily cadence and caches it, so your page never waits on anyone else's server. The widget then fetches that cached payload and renders it in the visitor's browser. That is a client-side render, which is exactly why this route is not an SEO play: the review text is not in the HTML the crawler receives, and as Step 1 established, self-serving reviews would not earn star snippets even if it were. What it buys you is a block that loads from your own origin, costs nothing per month, and cannot be emptied by a vendor's outage.

The second difference is visual, and it is the one clients notice. Every visual constant in the renderer, text colour, card surface, star colour, corner radius, border, shadow, card width and gap, reads a value that your page declares. Declare nothing and it renders in Google's own palette. Declare your design tokens and it disappears into the site. Ben's Curtains, a curtain and blind maker running an Australian and a New Zealand site, and LouvreSky, an Auckland louvre roof company with city landing pages, run the same app on sites that look nothing like each other, with nothing forked and no CSS overridden to force it.

Two honest constraints. Card selection can be filtered, by minimum rating, by minimum text length, or to reviews that came with a customer photo, but the headline score and the total count above the cards are always the real unfiltered Google figures. We will not print an average Google does not agree with, and neither should any tool you evaluate. And because the refresh runs on a schedule rather than on every page view, a review left this morning appears within the day, not within the second.

• Check whether your platform has a first-party renderer before paying a vendor for one; if it does, the monthly fee buys you nothing.
• A client-side render is a conversion asset, not a ranking asset, whoever ships it.
• Any tool that lets you filter the headline average, rather than just the visible cards, is one to walk away from.

Step 6: Put the reviews where the decision actually happens

Whichever route you took, placement decides whether it earns anything. The most common mistake is a dedicated testimonials page, which is the page nobody visits. Review blocks work where a visitor is already weighing up whether to contact you, which on most service sites means immediately above the enquiry form, and on most product sites means next to the price.

Put the summary bar, the star average and the review count high on the homepage, ideally within the first screen and a half, because that number is doing the work of an entire paragraph of copy. Put two or three full review cards further down, just before the call to action. A business with multiple trades or branches should put the relevant reviews on the relevant page rather than pooling everything on the homepage.

Photographs are the part most businesses skip and the part that persuades. A customer's own picture of the finished curtains, the laid floor or the completed roof does something a studio shot cannot, because it obviously was not art directed by you. If your tool can prioritise or require reviews that arrived with a photo, turn that on.

Finally, make the block bidirectional. A link from the summary bar straight into Google's own review form is the difference between a happy customer who meant to leave a review and one who did, and it costs one line of markup.

• Place the summary bar in the first screen and a half, and the cards immediately before the enquiry form.
• Prefer reviews carrying customer photographs; they outperform text-only cards by a wide margin.
• Link the block into Google's review form so the embed also feeds the supply it depends on.

Step 7: Check what the embed did to the page before you call it done

Run the page through PageSpeed Insights before you install and again afterwards, and keep both results. You are looking for three things: whether Largest Contentful Paint moved, whether Cumulative Layout Shift appeared, and how many new third-party requests the page now makes. A reviews block that lands late and pushes the page content down is a classic CLS source, and it is fixable by reserving the block's height in CSS rather than by removing the block.

Open Chrome DevTools, go to the Network tab, filter to third party, and count. A well-behaved embed adds one script and one data request. Several popular widgets add six or more, including fonts and analytics that have nothing to do with your reviews and everything to do with the vendor's own product telemetry. If the number surprises you, that is a finding, and it belongs in the decision you made in Step 2.

Then check the block at 390 pixels wide, the practical iPhone baseline, and confirm there is no horizontal scroll. Check it with a screen reader or at least with images disabled, which is where the screenshot route reveals itself as an empty rectangle. Last, click the link through to your live Google listing and confirm the number on your site matches the number Google shows. If it does not, and you took route one, you have just proved the point of Step 3.

• Capture PageSpeed Insights before and after; an embed that costs you 0.4 seconds of LCP is a trade, not a freebie.
• Count third-party requests in DevTools; one script and one data call is reasonable, six is a vendor's telemetry riding along.
• Test at 390 pixels and with images off, then verify your displayed rating against the live listing.

Frequently asked questions

Will embedding Google reviews improve my SEO or get me star ratings in search results? No. Google states that when the entity being reviewed controls the reviews about itself, including through an embedded Google Business reviews widget, its LocalBusiness and Organization pages are ineligible for the star review feature. Embed reviews to convert the visitors you already have, not to rank.

Can I show only my five-star reviews? You can filter which cards appear in most tools, and there are defensible reasons to do so, such as dropping one-word ratings that read as filler. What you must not do is filter the headline average. If your real Google score is 4.6, the number above the cards has to say 4.6, or you are publishing a figure your own listing contradicts in one click.

How much does a Google reviews widget cost? Manual screenshots cost nothing but decay. Third-party widgets run roughly USD 5 to USD 20 a month for small business traffic, which is USD 60 to USD 240 a year indefinitely, with free tiers capped low enough to be a trial rather than a plan. A platform-native renderer costs whatever your platform charges for it, which on WebForger is nothing.

How often do embedded reviews refresh? Third-party widgets and platform renderers both work from a cached copy that refreshes on a schedule, typically daily, rather than fetching live on every page view, because Google's data is not free to query and nobody wants the page waiting on it. A review left this morning normally appears the same day. Screenshots never refresh.

Do I need a Google API key or a developer to do this? Not for routes one or three. Route one needs a screenshot tool, and route three needs your platform to support it and your Place ID. Most third-party widgets also handle the data access for you, which is a real part of what the monthly fee buys.

The short version

Claim the listing, count your reviews, and only then choose a route. Screenshots are free and stale. A vendor widget is USD 60 to USD 240 a year and puts a third party's script on the page where people decide whether to call you. A platform-native renderer costs nothing extra if your platform has one, loads from your own domain, and takes its colours from your design rather than Google's. None of the three will win you a star rating in search, and all three can meaningfully lift the number of visitors who pick up the phone, if you put them above the form instead of on a testimonials page nobody opens.

If your site is on a platform that leaves you renting a widget for something this basic, that is worth knowing about the platform. Send us your current site and we will show you the same reviews rendered in your own colours, on a build where the app is included rather than billed.