When no language or region from the hreflang tag matches a user’s browser setting, the x-default tag tells Google what page to take users to. Setting a default is best practice for when no languages match.

The app lets you set one domain as the default in Settings > Domains, which uses a value of hreflang="x-default". Alternatively, no default can be set by disabling x-default in Settings. Individual pages cannot be changed to a default different to the store settings.

One of the main reasons we made this app is to solve the problem of different content between stores!

If an unpublished product page was included in the hreflang tags on the same SKU of another store, it would generate a 404. That’s not good for SEO.

Any product, collection, page, article, or blog not live will not be linked to via other hreflang tags. Additionally, items marked with Shopify’s seo.hidden metafield (noindex) are automatically excluded from hreflang tags, since pages hidden from search engines shouldn’t appear in hreflang annotations.

If you have a unique type of content on one store but no other, you can add it as a mapping (with it as the only entry) so that it has hreflang tags that point to itself. Some experts may argue this is unnecessary. Without any mapping, there are no tags on the page. If the store has multiple languages, adding the store as the only mapping adds the correct hreflang tags to it.

Pagination is a sequence of numbers for content. In ecommerce, it’s primarily a usability feature that lets people navigate between multiple pages of a product category. The alternative is to have all paginated products displayed on one page, which can cause confusion and excessive load times.

Shopify often paginates collection and blog content. How should you best handle hreflang tags for these pages to maximize your international SEO?

Shopify canonicalizes paginated URLs to themselves. The page /collections/all?page=2 will have a canonical tag of:

<link rel="canonical" href="https://example.com/collections/all?page=2">

Google’s John Mueller said hreflang tags must agree with that canonical:

Make sure any rel=canonical you specify matches one of the URLs you use for the hreflang pairs. If the specified canonical URL is not a part of the hreflang pairs, then the hreflang markup will be ignored.

The app follows this guidance by carrying the page number through every hreflang tag on paginated pages. Each paginated page declares itself and the matching paginated page in every other language or store:

<link rel="canonical" href="https://example.com/collections/all?page=2">
<link rel="alternate" href="https://example.com/collections/all?page=2" hreflang="en">
<link rel="alternate" href="https://example.com/collections/alle?page=2" hreflang="de-de">

Every paginated page gets a valid self-referencing hreflang tag that matches its canonical, along with reciprocal return tags on each alternate version. This is the structure Google expects.

What if the stores don’t match page for page? An English store may show products D, E, and F on page 2 while the German store shows F, G, and H, and a smaller store may have no page 2 at all. Neither breaks the markup. Hreflang on a category page tells Google “this is the equivalent category for this language or region”. It doesn’t require the two pages to list identical products, any more than translated pages have identical text. Shopify also never returns a 404 for a paginated URL. A page beyond the last simply renders with no products and still emits its canonical and reciprocal hreflang tags, so the tag pairs stay valid. At worst, Google quietly ignores a pair it can’t use. There is no penalty.

On the first page, the canonical drops the page parameter, so hreflang tags do too. The page /collections/all?page=1 has the following canonical and hreflang tags:

<link rel="canonical" href="https://example.com/collections/all">
<link rel="alternate" href="https://example.com/collections/all" hreflang="en">
<link rel="alternate" href="https://example.com/collections/alle" hreflang="de-de">

Shopify canonicalizes product variant URLs to the product URL. The page /products/product-name?variant=1234567890 will have a canonical tag of:

<link rel="canonical" href="https://example.com/products/product-name">

Variants URLs have hreflang tags that point to the primary URL, not the variant. This is called a “non-self referencing canoncial tag”. Following Google’s advice revealed in the FAQ “How is paginated content handled?“, this is done in order that hreflang tags match the rel=canonical link.

When auto-mapping products, the app expects your chosen match value—SKU, barcode, or a custom metafield—to be unique within each store. It links one product per store together into a single set of hreflang tags.

If it finds duplicates, such as two products in the same store sharing one SKU, it does not guess. It will not pick the first one it sees, and it will never generate multiple or conflicting hreflang tags from the ambiguity. Instead, it first tries to break the tie using the product’s URL handle. If exactly one of the duplicates has a URL handle matching the product being mapped, that one is mapped.

If the handle still can’t tell them apart, the app skips those products and leaves them for you to map manually. You’ll see an entry in the app’s Activity log, for example: "Couldn't auto-map [product] — sku [value] matched multiple items and the handle didn't uniquely resolve. Map manually if needed."

If a product is duplicated such that it shares the same SKU, barcode, or custom metafield of another, an existing mapping of the original product is unaffected.

You can map any flagged items by hand the same way you would in a spreadsheet, or switch your match method in Settings if another field is more reliably unique across your catalog.

New products and collections and their titles or URL changes sync almost instantly via webhooks.

Pages, blogs, and articles update daily by itself or instantly when you sync from the dashboard. Shopify does not provide webhooks for these content types, so this daily or manual sync is needed to pick up changes.

If your Weglot domains are configured as sub-folders with standard region and language folder names, the app will work.

If you have unique domains or sub-domains configured in Weglot, you cannot have perfect hreflang tags between Shopify accounts. Weglot’s domain configurations are a black box. They do not use Shopify’s APIs nor do they make the data easily available. A third-party app has no idea what these settings are in Weglot.

The alternative to achieve perfect hreflang tags for good SEO is setting up your domains in Shopify Markets or open new Shopify accounts. The domain and regions can be resynced through the app to the correct settings. Shopify makes the language, regions, and domains available.

Semrush and ahrefs always give an error of “Conflicting hreflang and rel=canonical” or “No self-referencing hreflang” for well-implemented hreflang tags in Shopify.

These messages are wrong with SEO best practices for hreflang tags. The errors occur because pages that are not the canonicalized version have hreflang tags that point to the canonicalized version.

This is the correct way to set up hreflang tags as Google’s John Mueller famously advised:

Make sure any rel=canonical you specify matches one of the URLs you use for the hreflang pairs. If the specified canonical URL is not a part of the hreflang pairs, then the hreflang markup will be ignored.

You may see a fluctuation of visibility report in Semrush or performance metrics in Google Search Console. This can happen the day the app is implemented as Google begins to correctly shuffle traffic between stores.

There are plenty of tools that will check tags on a page-by-page basis to confirm the errors reported in Semrush or ahrefs.com are false. The most reliable, comprehensive way to double-check your hreflang tags is with Screaming Frog. You can run a crawl of one of your stores then click the “Hreflang” tab at the top to see your tags.