This is a repost. Originally published on the Arvio blog: https://arvio.a.xyz/blog/shopify-products-not-showing-on-google. The canonical URL points back there.
Key takeaways
In one line: we ran the entire standard checklist against 28 real storefronts, and it passed on every store where it could be run. That will not tell you why your products are missing — for that you need your own Search Console, not 28 strangers' — but it does tell you which afternoon not to spend, and it points at the three things that actually varied.
- We measured the five things every article on this question tells you to check — canonical tag,
noindex,robots.txt, sitemap, structured data — on 26 live Shopify product pages. All five were already correct on all 26. - That makes the standard checklist a description of what Shopify does for you out of the box, not a diagnosis of your problem. 25 of 27 stores were running Shopify's default
robots.txt, unedited, down to the same 50Disallowlines. - What did vary is content. In 13 of 26 stores the product page's meta description is the opening of the product description, copied verbatim. 11 of 26 stop at 317–320 characters — roughly double what a search result displays.
- Product pages are heavy: median 443 KB of HTML, three of 26 over a megabyte, the largest 2.28 MB.
-
All 27 stores whose sitemap we could read are also serving
sitemap_agentic_discovery.xml, pointing at an/agents.mdfile and a live agent-commerce endpoint. Discovery is no longer only Google's, and nobody had to switch that on.
Published 25 August 2026 by Adot Technologies Inc, the team behind Arvio. Read-only measurements taken from public storefront endpoints on 25 August 2026 — one product page per store, no app installed, no admin access. We deliberately did not query Google itself; see the method section for why that matters to how you read this.
The quick version
-
Stop checking the checklist. On a stock Shopify storefront, canonical,
noindex,robots.txt, sitemap and product structured data are correct before you touch anything. We verified all five on 26 stores and found zero faults. - Work out which "not showing" you mean. Not in organic search results, not in the Shopping tab, and not yet crawled are three different problems, and only one of them is SEO.
- Then look at the page. The things that differ store to store are content-level: what the description says, whether there is one, whether the page is 400 KB or 2 MB, and whether the product can even be bought.
Three different problems called "not showing on Google"
Work out which one you have before any of the above. They are not related.
Not in organic search results. The blue links. This is the one the checklists are about, and on the evidence in the next section, on Shopify it is rarely a technical fault.
Not in the Shopping tab or product listings. That is Google Merchant Center, a separate system with its own feed and its own approval process. Nothing in your theme, sitemap or robots file affects it. If this is your problem, the answer is in Merchant Center's item-level diagnostics, not in your store's SEO.
Not yet crawled. A new URL is not a missing URL. Discovery takes time, and it is uneven page to page.
That last one happened to us this month, on our own site. Four articles published within days of each other, one sitemap, identical templates: three were in Google's index within the week and one was not. Same setup, same submission. Nothing distinguished the missing one except that Google had not got to it. If we had gone looking for a technical cause, we would have found something plausible and fixed the wrong thing.
How to tell which one is yours
In order, because each step rules out the ones after it:
- Search your own store's exact product title in quotes. (If your titles came from a supplier and read like everyone else's, this step will return their pages, not yours, and tells you nothing — search your store name plus the product instead, or skip to step 2.) If the page appears, you are indexed, and your problem is ranking or click-through rather than discovery — which is a different job. The most common quick win there is the snippet itself: see what your meta descriptions actually say, since on half the stores we read it is the opening of the product body, cut mid-sentence.
- Check whether you mean Shopping. If the missing thing is the product carousel with prices, the answer is in Merchant Center's item-level diagnostics, not in an SEO guide.
- Open the product page and read the description. If it is empty, thin, or identical to the supplier's, nothing else you do will matter.
- Read the page's meta description. If it is 320 characters of the product body, decide whether the visible first 150 are the ones you would have chosen.
- Check whether the product can be bought. A sold-out page still gets submitted, still gets crawled, and still absorbs whatever traffic reaches it.
-
Only then open
robots.txtand view-source for a canonical. On the evidence in the next section, this step is where you confirm nothing is wrong, not where you find the problem.
What we measured
One product page per store, taken as the first product the public catalogue returned, across 26 storefronts.
| Check | Stores passing |
|---|---|
| Canonical tag present | 26 / 26 |
| Canonical points to itself | 26 / 26 |
No noindex on the page |
26 / 26 |
| Product structured data present | 26 / 26 (25 as JSON-LD, 1 as microdata) |
robots.txt served |
27 / 28 |
robots.txt names a sitemap |
27 / 27 |
sitemap.xml served |
27 / 28 |
| Product sitemap matches the catalogue | every store we could count, to within one URL |
Read the denominators. The first four rows are one product page each on the 26 stores that served one. The robots.txt and sitemap.xml rows are 28 stores, and one of those 28 is the reason they say 27 rather than 28: it answered neither request, while serving a 3,250-product catalogue perfectly well from the same domain. We could not check those two things on that store, so it is not a pass and we are not going to round it into one — and a storefront that will not answer robots.txt or hand over a sitemap is exactly the kind of discovery problem this article is about. One store in 28.
With that said: on every store where a check could be run, it passed. Not one canonical pointing at the wrong URL, not one stray noindex, not one missing product schema. If your products are not showing, the reason is almost certainly not on that list.
The default nobody edits
Every guide starts you on robots.txt. 25 of the 27 stores were running Shopify's default file unedited — the same 50 Disallow lines, in the same order. Two stores had customised theirs, at 144 and 159 lines. Every single one named a sitemap.
Shopify's default blocks the things that should be blocked: cart, checkout, account pages, internal search, the URLs that generate infinite filter combinations. Products are not in it. Unless you or an app deliberately edited that file, it is not what is keeping your products out.
The description nobody wrote
Everything above was the same on every store. This is not.
We compared each product page's meta description against the product's own description text:
| Stores | |
|---|---|
| Meta description present | 26 / 26 |
| Meta description is the opening of the product description, verbatim | 13 / 26 |
| Length 317–320 characters | 11 / 26 |
| Length over 160 characters | 18 / 26 |
| Median length | 218 characters |
| Shortest / longest | 23 / 1,811 characters |
The 317–320 cluster is the tell, and it is worth being precise about what it is. Every one of those eleven stores has a meta description that is the opening of its own product body, cut off. Their product descriptions are nothing like each other in length — 427, 579, 636, 942, 2,088, 2,250 characters — and yet all eleven land within three characters of the same ceiling. Different catalogues, different themes, different countries, one number.
That is a 320-character truncation being applied for them, on a field none of them filled in. From outside the store we cannot see which piece of code emits it, and we are not going to guess between Shopify's default and a theme convention. What we can say is that it is not a decision any of these merchants made: the text is their product copy, the cut is somebody else's, and the join is mid-sentence.
Set your own SEO description on a product and this stops happening — that field, when it is filled, is what ships. It is on the product page in the admin, under Search engine listing, behind the Edit link; empty by default on every product.
A search result shows roughly the first 150–160 characters. For the stores sitting at the automatic ceiling that means about half of what is there never appears; at the median length, closer to a third. And whatever does appear is whatever the product copy happened to open with.
We are not claiming this is a ranking factor. The point is simpler: the sentence that decides whether someone clicks is, in exactly half these stores (13 of 26), a sentence nobody chose.
The weight of the page
Median product page: 443 KB of HTML. Three of the 26 were over a megabyte, the largest at 2.28 MB. That is the document alone, before images, fonts or scripts finish loading.
For scale: the visible text on those pages runs to a few thousand characters. The other 400-odd KB is theme and app markup, and it ships on every product page in the catalogue. We did not measure rendering, Core Web Vitals or crawl budget, so this is not a ranking prediction — it is a place to look first if your store is slow, and a reason to check what each installed app adds to a page you serve ten thousand times.
Empty pages and unbuyable products
Across the same 28 storefronts, pooled: 1.0% of published products had no description at all (291 of 29,009), and 6.6% had no image (1,910).
Do not read those as rates, because they are not. Only 8 of the 28 stores have a single product without a description, and 276 of the 291 — 95% — are one store; the next worst has four. Missing images are worse still: two stores supply 1,904 of the 1,910. This is not a condition most catalogues have a little of. It is a condition a few catalogues have a lot of, and the pooled percentage hides exactly that.
So the honest version is: empty product pages are rare across stores and severe within the stores that have them. If you are one of those stores you already half-know it, and it outranks everything else in this article.
And from the same crawl: the median store has 3.9% of its catalogue entirely out of stock and still published, with nine of 28 stores above 10%. We followed one of those products through and it returned HTTP 200, displayed "Sold out", and was still listed in the store's product sitemap — still being submitted to search engines as a page worth having. We wrote that measurement up separately in how much of your Shopify catalogue is unbuyable right now.
An empty product page is a page with nothing to rank. That is the most boring possible explanation for "not showing on Google", and it is the only one on this list you can act on today. We measured how often the condition turns up, not how often it is the cause — but it turns up, and nothing on the checklist did.
Where to start, if that is you. Not with the 400 products. Open your product sitemap (/sitemap_products_1.xml), take the twenty products you would most like to be found, and read those pages as a stranger: is there text a search engine could quote back, or is it a title, a price and a supplier's spec table? Fix those twenty, watch them for a month, and you will know whether thin content is your problem before you commit to the other 380. Everything else in this article is diagnosis; this is the part that is work.
Discovery is not only Google's any more
Every one of the 27 stores whose sitemap we could read was already serving an agent-discovery sitemap, and all 28 answered on an /agents.md instruction file and a live MCP endpoint — none of which any of these merchants switched on. It is a real finding but it is not this article's question, and we have not measured whether any of it carries traffic. It is written up separately in every Shopify store we read is already serving an agent API.
Method
Sample. 28 live Shopify storefronts, of which 26 served a product page we could read, on 25 August 2026. The frame is a convenience sample of storefront domains from stores that have installed our app at some point — not random, not representative of Shopify overall, skewed small and young. Excluded before results were examined: storefronts without a public catalogue (password-protected, closed, or unpaid), invalid JSON responses, and any store with fewer than 10 published products.
Collection. Read-only requests to public endpoints at one request per second with a self-identifying user agent: the products endpoint, robots.txt, sitemap.xml, the first product sitemap, and one product page per store. No admin access, no authenticated session, no writes.
Two corrections we had to make mid-measurement, both worth knowing if you repeat this:
- Our first pass stopped paging the catalogue when a page returned fewer items than requested. That endpoint can return a short page with more pages behind it — one store read as 238 products instead of 3,100. Stop on an empty page, never on a short one.
- Our first structured-data check only scanned the first 200,000 characters of each page and reported 10 of 26 stores with product markup. Since the median page is 443 KB, that check was measuring page length. Re-run against the whole document, the real figure is 26 of 26. The first number was wrong and we are documenting it here rather than quietly publishing the second one.
What we did not do: ask Google anything. No index checks, no rank queries, no Search Console access to stores we do not own. So this is a measurement of the prerequisites for being indexed, not of whether any of these stores is indexed. Nothing at this sample size can support the second claim, ours included.
Limits, in one place
- Whether any of these stores' products actually appear in search. Not measured, by design. See above.
- Rendered pages. We read the HTML the server sent. Anything a theme or app injects after JavaScript runs is invisible to our measurement — and to any crawler that does not execute scripts.
- One page per store. The product we checked is the first one the catalogue returned. A store can have a correct first product and a broken thousandth one.
- Cause and effect anywhere. We have not tested that a longer meta description, a heavier page or an out-of-stock product changes any outcome. We measured what is on the pages.
- The agent endpoints beyond existence. We confirmed the discovery documents respond. We did not transact, and we have no data on whether anything reaches stores through them.
Whether the checklist ever mattered on Shopify, or whether it was inherited from WordPress-era SEO where those files really were hand-edited and really did break. Our 26 stores can only tell you it is not broken now. A version of this measurement on 500 stores, run monthly, would settle it — and would be the first evidence anyone has published either way.
FAQ
Why are my Shopify products not showing on Google?
On the evidence here, rarely because of robots.txt, a canonical tag, noindex or a missing sitemap — every one of those was correct on all 26 stores we measured. The likelier causes are that the page has not been crawled yet, that you mean Google Shopping rather than search, or that the page is empty enough that there is nothing to show.
How long does it take Google to index a new Shopify product?
There is no fixed answer, and it is uneven within one site. On our own site this month, of four pages published days apart with one sitemap and identical templates, three were indexed within the week and one was not. A page missing after a few days is not evidence of a fault.
Does Shopify block Google in robots.txt?
No. 25 of the 27 stores we read were serving Shopify's default file unedited, and it blocks cart, checkout, account and internal search URLs — not products. All 27 named a sitemap.
Do I need to submit my Shopify sitemap to Google?
Your store already generates and links one, and it matched the published catalogue to within one URL on every store we could count. Submitting it in Search Console does no harm and gives you the coverage report, which is the actual reason to do it.
Does Shopify add product structured data automatically?
On the 26 pages we read, product markup was present on all of them — 25 as JSON-LD and one as microdata. Whether it is complete enough for rich results is a separate question we did not test.
Why is my meta description not showing in search results?
Often because it is too long. Half the stores we measured had one over 160 characters, and 11 of 26 sat at 317–320 — the point where an automatic fill stops. A search result shows roughly the first 150–160 characters, so everything after that is invisible where it counts.
Should out of stock products be removed from my sitemap?
Not if the product is coming back — you would be throwing away the page's history. The thing worth checking is not the sitemap entry. It is whether you are still sending people to a page that cannot take an order.
What is the agents.md file on my Shopify store?
A file the platform publishes describing how AI shopping agents should interact with your store, announced through a second sitemap called sitemap_agentic_discovery.xml. Every store we checked had one, without anyone enabling it. It points at a discovery document and an agent-commerce endpoint that were live on all three stores we tested.
Related: How much of your Shopify catalogue is unbuyable right now — the same crawl, counting products where every variant is unavailable.
Related: Your SEO app found 400 missing meta descriptions. Now what? — what to do once you know which pages are thin.
Written by Adot Technologies Inc, the team behind Arvio: AI Store Operator — ask what is missing across your catalogue, review every edit before it goes live, and undo anything.
Originally published at https://arvio.a.xyz/blog/shopify-products-not-showing-on-google. More Shopify bulk-editing writeups are on the Arvio blog.
Top comments (0)