Discussion starter: an ecommerce SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- eCommerce SEO
- Indexing
- Shopify SEO
On Shopify collections, which filter combinations do you let Google index, and how do you enforce it with the platform’s URL constraints?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
On most Shopify stores I let Google index almost none of the filter-parameter URLs themselves. The combinations that deserve to rank are the ones with their own search demand, enough products to be a real choice, and inventory that stays stable. Those I turn into proper collections with their own handle, title, description and internal links. Every other filtered state stays out of the index: a server-rendered noindex for single-filter URLs, and Shopify’s default robots.txt rules for multi-filter and sorted URLs.
The reason is the platform. Shopify filters are query strings on a fixed
/collections/path, the sitemap is generated for you, and your control over the head comes through theme Liquid. You cannot give a filter state a clean URL, unique on-page copy or a sitemap entry without workarounds. A dedicated collection gets all three natively, so “promote, do not expose” is both the cleaner SEO outcome and the easier one to maintain.What Shopify generates and what it already blocks
Shopify’s storefront filtering documentation describes parameters such as
filter.p.product_type=shoesand variant options such asfilter.v.option.colour=black(the last part follows your option name). “Filters are applied withANDlogic, and filter values withOR.” Multiple values can be sent as a comma-separated list or as repeated parameters, and a store can create up to 25 filters. Legacy tag filtering adds another pattern:/collections/boots/black, with+joining multiple tags.Shopify’s default robots.txt, which you can see on most stores that have not customised it, already handles part of this. It includes rules such as:
Disallow: /collections/*sort_by*(any sorted collection URL)Disallow: /collections/*+*plus encoded variants (multi-tag combinations)Disallow: */collections/*filter*&*filter*(two or more filter parameters)
What the defaults leave crawlable matters more. A URL with one filter parameter is open, including one parameter carrying several comma-separated values. A single legacy tag URL is open. If your theme sends multiple values as repeated parameters, those URLs are blocked; if it sends them comma-separated, they are not. The same filter state can therefore be crawlable or blocked depending on theme code.
Hypothetical example:
A fashion store has 150 collections and filters for colour (12 values), material (6), size (14), brand (25) and availability (2). Counting only single-value, single-filter URLs, that is 150 × 59 = 8,850 crawlable URLs before comma-separated value combinations and price ranges. Keyword research shows real demand for around 40 of those states, such as “black leather boots” or “linen shirts”. The other 8,800 or so are useful to shoppers and worthless as search landing pages.
Whether Google finds those URLs also depends on the theme and any filter app. Facets rendered as crawlable
<a href>links expose every value; facets rendered as form inputs expose fewer. Third-party filter apps often use their own parameter names, which Shopify’s default robots rules do not cover. Check what your storefront actually outputs before deciding how aggressive you need to be.Which combinations earn a place in the index
Filter type Usual verdict Why Colour, material, style within a core collection Promote the ones with demand Shoppers search “black leather boots”; the modifier changes the product set meaningfully Brand within a category Promote for brand-led categories “Brand + category” queries are common; build brand collections rather than relying on /collections/vendors?q=pagesTwo-attribute combinations Only with clear demand and product depth Demand exists for some (“women’s waterproof hiking boots”); most combinations are too thin Size Usually not; exceptions for fit types “Size 9” pages churn with stock. “Wide fit” or “plus size” can be a real category Price range, availability, sort order Never No distinct intent, constantly changing product sets For each candidate, I want evidence of demand from more than one place: keyword data, Search Console queries already reaching the parent collection with that modifier, internal site search terms, and Google Ads search terms if the store runs Shopping or Search campaigns. Then I check product depth across the year, not just today. A combination that is strong in winter and empty in summer needs a plan for the empty months (keep it live with alternatives, or accept it will underperform), not a page that turns into a thin shell.
Filter combination↓Distinct search demand, not already served by the parent collection? If no: keep out of the index↓Enough products all year, and commercially worth a page? If no: keep out of the index↓Can you give it a unique title, H1, short intro and internal links? If no: not yet↓Promote to a dedicated automated collectionHow to enforce it within Shopify’s constraints
1. Promote demand-backed combinations to real collections
Create an automated collection whose conditions reproduce the filter state, using conditions such as product tag, type, vendor, category or supported metafields. It gets its own handle (for example
/collections/black-leather-boots), its own SEO title and description, a place in Shopify’s generated sitemap, and a canonical that points to itself. This is the only route where every signal is under your control.A collection with no links is barely better than a filter URL, so link it. Add a “Shop by” block on the parent collection and link from navigation, buying guides and related collections. Some stores go further and make the matching filter value link to the promoted collection. That can work, but it changes how the filter UI behaves, so test it with real shoppers before rolling it out.
2. Noindex every filtered state in the theme
Liquid cannot easily read the query string, but the collection’s filter objects expose which values are active. A snippet like this in
theme.liquid, inside the head, is a starting point to adapt and test on an unpublished theme first:{%- liquid assign filtered = false if template.name == 'collection' if current_tags.size > 0 assign filtered = true endif for filter in collection.filters if filter.active_values.size > 0 assign filtered = true endif if filter.min_value.value != blank or filter.max_value.value != blank assign filtered = true endif endfor endif -%} {%- if filtered -%} <meta name="robots" content="noindex"> {%- endif -%}Confirm the tag appears in the raw HTML response, not only after scripts run. Google’s JavaScript SEO documentation warns that robots directives changed by JavaScript may not behave as expected (a noindex in the original HTML can stop rendering altogether), and many filter apps update the page client-side. Put the directive in the server response. Check too that no SEO app is printing a second, conflicting robots tag. Handle
/collections/vendorsand/collections/typespages the same way if your theme links to them.3. Keep robots.txt for crawl control, not index control
Shopify lets you edit rules through a
robots.txt.liquidtemplate. The default multi-filter and sort rules are sensible, so leave them. Add rules only when something specific leaks: a filter app’s own parameters, or single-filter URLs at a scale that genuinely competes for crawl. Google’s crawl budget guide is aimed at sites with a million or more pages changing weekly, or 10,000+ pages changing daily, so most Shopify stores are dealing with index quality, not crawl budget.Watch for this:
Do not block URLs in robots.txt that are already indexed and that you want removed. A blocked URL cannot show Google its noindex, so it can linger, and linked blocked URLs can appear as “Indexed, though blocked by robots.txt”. Noindex first, wait for them to drop out, then block if crawl volume justifies it.
4. Do not rely on the canonical alone
Most themes print
{{ canonical_url }}, which Shopify’s Liquid reference defines only as “The canonical URL for the current page”. It does not document which parameters survive on filtered collection URLs, so load a filtered URL, view source, and compare it with the Google-selected canonical in URL Inspection. Even when the canonical points to the parent collection, it is a hint. A filtered page showing a different product set is not a duplicate, and Google may ignore the hint (“Duplicate, Google chose different canonical than user”) or index the filtered URL anyway. Use the canonical as support for the noindex, not as the control.Migrating a store that already has filter URLs in the index
- Export Search Console page data and isolate URLs containing
filter.or tag paths. Sort by clicks. - For filter URLs earning meaningful clicks or revenue, create the promoted collection first, link it internally, and only then noindex the filter version. Otherwise you trade a ranking URL for nothing.
- Deploy the noindex snippet for everything else.
- Watch the Page indexing report: filter URLs should move into “Excluded by ‘noindex’ tag”, and promoted collections should appear as indexed. Watch “Duplicate, Google chose different canonical than user” for signs Google prefers a filter URL over your new collection.
- Track the promoted collections on business metrics: organic sessions, add-to-cart rate and revenue. A page that ranks but converts worse than the parent is not a win.
A reasonable test is to promote the top 10 to 20 combinations first, compare their query coverage and revenue against the same filter URLs’ performance in the previous period, and expand only if the promoted pages earn impressions for modifier queries the parent collection was not ranking for. If they do not, the modifier demand was already being served, and the promotion list should be shorter.
A practical position
If this were my store, the filter UI would stay fully usable for shoppers, every filtered state would carry a server-rendered noindex, Shopify’s default robots rules would stay in place, and a short, evidence-backed list of combinations would become real collections with copy and links. That list would be reviewed each season against search demand and stock depth, because on Shopify the collection list, not the filter parameters, is where you decide what Google should rank.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.