How to Migrate a Client’s Reviews Without Losing Their Google Stars

Why do Google stars disappear when a store changes review app?

Because Google does not store a store’s stars. It reads them, on each crawl, from the Review and Product structured data the review app prints into the product page, and it checks that the rating in the markup matches the reviews a visitor can see. A migration breaks that in one of three ways: the markup disappears for a while, two apps print two different ratings for the same product, or the markup and the visible reviews stop agreeing.

Each of those is avoidable with ordering. The playbook below is written for an agency doing this on a client’s store, where the client expects the switch to be invisible to their customers and to Google, and where you are the one who answers for it if the stars go.

1. 01Markup gap: the old app is removed before the new one prints anything.
2. 02Double markup: both apps print an AggregateRating for the same product.
3. 03Mismatch: the markup says 412 reviews and the page shows 38, or none.

Step one: how do you export the reviews from the old app?

Export first, before you change anything else, and while the client’s subscription is still active. Most review apps export reviews as a CSV file from the reviews or settings area of their admin, and larger exports are often emailed rather than downloaded. A few specifics, read on each vendor’s help centre on 24 September 2026: Judge.me exports from Reviews, using the Export dropdown at the top right, as CSV with up to 100,000 reviews per file, and videos and SKUs are not in the file. Loox exports from Manage Reviews, under the three-dot menu, with product and store reviews as separate CSV files. Junip exports from its Reviews page with the Export button at the top right, and emails the CSV when it is ready. For Yotpo, Okendo and Stamped, export your reviews as CSV from the app’s settings, and if you cannot find a full export, ask the vendor’s support in writing for one.

Whatever the app, keep the original file untouched and work on a copy. Emailed export links can expire within hours, so download promptly.

1. 01Export product reviews and store reviews, if the app keeps them separately.
2. 02Check that the file has body, rating, author, date and a product identifier.
3. 03Photos and videos: confirm whether the file carries links, and download the media while the account is live.
4. 04Reconcile the row count against the old app’s dashboard before relying on it.

Step two: how do you import without orphaning reviews?

Match every review to the right product before the import, not after. Apps identify products differently, by Shopify product ID, by handle, or by SKU, and a column that does not line up either orphans reviews or lands them on the wrong product. Pick the one key both apps agree on, spot-check ten products by hand, and hold back reviews for products that have since been deleted or merged.

Keep the original author names and dates. A review dated today that was written two years ago misrepresents the store’s history, and resetting every date also throws away the recency pattern buyers use to judge a product.

If the new app is BeyondReviews, the import is part of setup rather than a separate project: the client’s past reviews come across and count toward their rating from day one. On Studio and above, those existing reviews are also read and answered at signup, with every reply drafted in the store’s voice for someone to approve, and Atelier and Maison can add top-up packs for larger back catalogues. Maison includes a white-glove migration.

Step three: how do you keep Review and Product schema continuous?

The goal is simple to state: at every moment of the migration, each product page carries exactly one set of review markup, and it agrees with the reviews on the page. Before you start, open three or four representative product pages, view the source, and note what is printing Review and AggregateRating today: the review app, the theme itself, or both. Many themes print their own Product schema, and some include a rating block that reads from an app’s metafields.

After the import, check that the new app’s rating and count for those same products match the old app’s, allowing for any reviews you deliberately held back. If they do not match, stop and fix the mapping before any markup changes hands. Run the pages through Google’s Rich Results Test, or paste the JSON-LD into our free schema validator at /tools/schema-validator/, and keep a note of the before and after for the client.

How do you avoid two AggregateRatings during the overlap?

Run both apps briefly, but never let both print markup at once. The overlap is there so the product pages never go blank; it is not a reason for Google to see two ratings for one product. Two apps each printing an AggregateRating hand Google two different answers to the same question, and disagreement between markup and page is the most common reason stars are pulled.

In practice: install the new app, import, and verify with its widgets hidden or on an unpublished theme copy. Turn off the old app’s rich snippet or schema setting if it has one. Then make the switch in a single step, publishing the theme with the new widgets and schema at the moment the old ones come out.

BeyondReviews ships schema dedup on every plan, including Free, built so that only one star rating per page reaches Google even when a theme block or an old app is still printing its own. It is still worth viewing the source after the switch; a dedup is a safety net, not a reason to skip the check.

1. 01Do the new install and import on an unpublished copy of the theme.
2. 02Switch off the old app’s rich snippet or schema output.
3. 03Publish the new widgets and remove the old ones in the same change.
4. 04View source afterwards and count the AggregateRating blocks per page: one.

When should you swap the widgets and uninstall the old app?

Swap in a quiet hour, not on a launch day, and not during a sale when the client will be too busy to check anything. Replace every surface the old app touched: the product page rating and reviews block, collection page stars, any home page carousel, and the review request emails. Theme app blocks and app embeds come out in the theme editor; older installs sometimes left snippets in the theme code, so search the theme files for the old app’s name after you remove it.

Uninstall the old app only after you have confirmed the new widgets, the new markup and the review requests on real pages. Keep the original export file and any downloaded media for at least a few months: it is the only complete copy if anything needs re-importing.

Do you need redirects if review page URLs change?

Only for URLs that actually change. Product URLs normally stay exactly as they are through a review app migration, and those are the ones that carry the rankings. What can change is a standalone reviews page, such as an all-reviews page under /pages/, or review pages hosted by the old app, and any links to them in emails, footers or ads.

List every URL the old app served, check which ones Google has indexed with a site: search or in Search Console, and add a 301 redirect from each old path to its nearest new equivalent using Shopify’s URL redirects. Point an old all-reviews page at the new one, not at the home page. Update internal links in the theme and email templates at the same time, so the redirects are a safety net rather than the main route.

What should you check in Search Console afterwards?

Check at the switch, then weekly for two to four weeks. In Search Console, use URL Inspection on a handful of product pages to request a recrawl and confirm Google sees the new review markup. Then watch the rich result reports for product and review snippets: the count of valid items should hold roughly steady, and any new errors or warnings usually point to a template that was missed.

Watch the Performance report too. Filter by the product pages you migrated and compare clicks and click-through rate against the weeks before the switch; a drop in click-through with steady impressions is the usual sign that the stars have gone from the results. Send the client a short note at the end of the window with the before and after, because a migration nobody noticed is exactly the outcome they paid for.

1. 01URL Inspection on a sample of product pages, with a recrawl request.
2. 02Product and review snippet reports: valid items holding steady, no new errors.
3. 03Performance: clicks and click-through rate on migrated pages, before and after.
4. 04Page indexing: old review page URLs showing as redirected, not as 404s.

Where does BeyondReviews fit in a client migration?

As the destination, where it suits the client. BeyondReviews does the display job on every plan including Free, photo and video reviews, every storefront widget and Google stars through Review and Product schema, and it treats keeping a store’s stars through the move as our problem rather than the client’s. It then goes further: Chorus counts the phrases customers repeat, approved phrases are printed into the product page’s own HTML where Google and AI answer engines read them, and every reply is drafted in the store’s voice for approval. Every plan exports everything, whenever the store likes, so the next migration, if there ever is one, is easy too.

If you move clients as part of your work, the partner programme at /partners/ pays 50% of subscription revenue for each referred client’s lifetime. Before a migration, the free AI readiness checker at /tools/ai-readiness-checker/ shows whether the client’s current reviews reach the HTML crawlers read, which is a useful baseline to show them afterwards.

Frequently asked questions