Skip to content

E-commerce SEO · guide 4 of 4

Filters and faceted navigation: keep the useful pages, stop the URL explosion

Shop filters can create millions of near-identical addresses. Learn which filter pages deserve to be in Google, which controls actually work, and how to check it in Search Console.

22 min readUpdated 9 Nov 2026

  • 8 questions to test yourself
  • 1 calculator
  • 13-item checklist
On this page11 sections

Your shop may be bigger than you think

You sell 400 products. In Google’s eyes, your shop might have four hundred thousand pages, or four hundred million. Nobody wrote them. Your filters did.

Every time a shopper ticks “size 38” or “red” and the address changes to something like /boots/?colour=red&size=38, your shop has made a new URL: one more page Google can find, request and try to understand. Google calls this faceted navigation, and Google Search Central has a whole page on managing it.

By the end you will have the number for your own shop, a short list of filter pages worth keeping, the controls that actually stop the rest, and a score showing where your setup stands. Along the way there is one fix that feels doubly safe and quietly backfires; see if you can spot it before we get there. Start with a guess.

Quick quiz

A shoe category has 5 filters (size, colour, brand, material, width) with 6 options each, and shoppers can tick as many as they like. Ignoring sort order, roughly how many filter combinations can it produce?

Pick one. Your first answer is the one that counts.

Count the addresses on one of your own categories

The quiz used made-up numbers. Here are yours. It takes a minute. Watch what the second line does when you add one option to a filter.

Try the numbers

The URL explosion counter

Pick your biggest category. Count its filters, the typical number of options in each, and the sort orders in the “Sort by” menu. Keep the top number in mind: you will meet it again in Search Console.

4 filters
6 options
4 sort orders
200 products

Addresses if shoppers can pick one option per filter

9,604 URLs

Addresses if they can tick up to two options per filter

937,024 URLs

One-pick addresses per product you sell

48 per product

A count of addresses that could exist, not a forecast of what Google will crawl. Allowing three or more ticks per filter, pagination or a different parameter order multiplies it further; your platform may already block some.

Now the test that decides whether that number is real for Google. Open the category, right-click one filter and see whether your browser offers “Copy link address”. Then tick it and watch the address bar.

If it is a link, or the address changes to something with a question mark, Google can follow it and your number is live: the rest of this guide is for you. If it is not a link and the address never changes, Google sees close to none of those combinations; read the FAQ on JavaScript filters, then the section on promoting the filters worth ranking.

Why so many addresses hurts, and when it does not

Google does not crawl a billion URLs on a small shop, and you are not penalised for having filters. The harm is quieter. Google’s faceted navigation documentation names two problems: overcrawling, where Googlebot spends its visits on filter combinations that add nothing new, and slower discovery, where your new products and updated pages wait longer to be crawled.

There is a third cost you can see in Search Console: thousands of near-duplicate pages that Google crawls and then declines to index. They show the same 20 boots in a different order or a slightly smaller set. Any links or attention they collect are spread across copies instead of going to your real category. How Google picks one copy to show, and when a redirect beats a canonical tag, is the subject of the guide on duplicate content and canonicals; this guide stays with the filters.

For context, Google’s guide to managing crawl budget says it is written mainly for very large sites: roughly a million or more unique pages that change weekly, or 10,000 or more that change daily. A 400-product shop is nowhere near that in real pages. Filters are how a small shop ends up with a large shop’s crawling problem.

Myth or fact?

Four things shop owners believe about filters

For each one: myth or fact?0 of 4 answered
  1. 1.Search Console has a setting where you tell Google which URL parameters to ignore.

  2. 2.A page blocked in robots.txt can still show up in Google’s results.

  3. 3.A canonical tag on every filter page stops Google crawling them.

  4. 4.Filters only matter for huge shops.

Which filter pages deserve to be found

A few filter pages are genuinely valuable. Someone typing “waterproof hiking boots women” wants exactly the page your filters can build. Most combinations are not: nobody searches for “red size 38 boots under €80 sorted by newest”.

A filter page earns a place in Google only when all four of these are true:

  • People really search for it. Check Search Console or a keyword tool, not your instinct.
  • It shows enough products to be a useful choice.
  • It stays stable from week to week.
  • It is meaningfully different from the category it belongs to.

Sorting, view switches (grid or list), “in stock only”, price sliders, items per page, session IDs and tracking codes almost never pass. Try six real cases: decide before you flip each card. One is a trick.

Test yourself

Think of your answer, then flip the card.0 of 6 flipped

Turn the winners into proper pages

When a filter passes the test, do not just let the parameter URL get indexed. Make it a page you would be proud of. That usually means a clean, permanent address, a title and heading written for the search (“Women’s waterproof hiking boots”), two or three sentences that help someone choose, and a link to it from the parent category and, if it matters enough, your menu.

On Shopify this is often a separate collection; on WooCommerce a subcategory or an attribute archive; on a custom shop, a route your developer adds. The filter stays in the sidebar for shoppers; the promoted page is what you want Google to find.

Write it the way you would write any category: the guide on e-commerce category pages covers the title, the intro that helps people choose and the links that make a category easy to reach. If the page can only be reached through the filter itself, the internal linking guide shows how to give it links from the pages Google already visits.

Before and after

Left as a filter

Address: /boots/?attr_feature=waterproof&gender=w&sort=popular

Title: Boots | Example Outdoor Shop

Heading: Boots

No text. The same page also exists as /boots/?gender=w&attr_feature=waterproof and with every sort order.

Promoted to a page

Address: /boots/womens-waterproof/

Title: Women’s waterproof hiking boots | Example Outdoor Shop

Heading: Women’s waterproof hiking boots

Intro: Membrane versus waxed leather, how to pick a size if you wear thick socks, and our three lightest pairs. Linked from Boots and from the Hiking menu.

  • One address instead of dozens, so every link and every click points at the same page.
  • The title and heading now match what people type, which tells Google and the shopper what the page is about.
  • Illustrative example: the shop and URLs are made up.

To see this decision made across a whole shop, read the Fietsgaard case study on our Case studies page. It is illustrative, with an invented three-store bike shop in Utrecht and modelled figures, not a client result. Its frame-size, motor and brand filters were plain links that created tens of thousands of thin addresses; the study follows 64 searched-for combinations becoming real landing pages, while deeper combinations lost their crawlable links so Google stopped finding new ones.

The controls, and what each one actually does

This is where most shops go wrong: they reach for the tag they have heard of without knowing what it controls. There are two separate questions. Can Google crawl the page (request it)? And will Google index it (keep it and show it in results)? Each tool answers only one of them. Test yourself on the fix promised at the start before you see the table.

Quick quiz

Here is the fix that feels doubly safe. Your filter pages carry noindex, but Google still crawls them heavily, so a colleague blocks them in robots.txt the next day. What happens to the filter URLs already in Google’s index?

Pick one. Your first answer is the one that counts.

Side by side

ControlStops crawling?Keeps it out of the index?Use it for
robots.txt Disallow ruleYesMostly; a linked URL can still be listed without contentFilter and sort parameters you never want in Google; Google’s first suggestion for them
noindexNo, Google must crawl to see itYes, once crawledA page you want crawlable but not listed. Weak for filters, because the crawling continues
Canonical tagNoUsually; it is a hint Google can overruleSort orders and near-duplicates, pointing at the main version
Filters without crawlable links (JavaScript, or after a #)Yes, Google generally ignores the part after #Yes, there is nothing to indexNew builds and redesigns; ask your developer
404 status for empty resultsStops repeat visits over timeYesAny combination with zero products
Summary of Google Search Central’s guidance on faceted navigation, robots.txt and canonicalization. Your platform may handle some of this already.

A sensible setup for most shops

You do not need every control. For most small and mid-sized shops, five steps cover it, and the order matters more than the steps. Before you open them, put them in order yourself.

Put it in order

Put the five steps of a filter clean-up in the right order

Use the arrows to move each item, then check.

  1. Promote them to clean pages
  2. Pick the few worth indexing
  3. Tidy what remains crawlable
  4. Block the rest from crawling
  5. List every filter parameter

Step by step

From messy filters to a clean set of pages

Step 1 of 5

List every filter parameter

Click through two or three categories, tick filters and sorts, and write down what appears in the address: colour=, size=, sort=, price=, view=. Also open yourshop.com/robots.txt and see what is already blocked. Shopify, for example, ships a default robots.txt that covers some sort and multi-filter addresses.

Monitoring: how to tell it is working

Remember the top number from your counter? Search Console is where you see how much of it Google has actually found. Three places tell you most of what you need. Write down today’s numbers so you have something to compare against in a month.

  • Indexing, then Pages: open the reasons pages are not indexed and click the examples. “Duplicate without user-selected canonical”, “Crawled – currently not indexed” and “Discovered – currently not indexed” full of ?colour= and ?sort= addresses are the classic sign of a filter problem. “Indexed, though blocked by robots.txt” means blocked URLs are still listed.
  • Settings, then Crawl stats: see how many requests Googlebot makes and open the example URLs. If most are filter combinations, that is crawling not spent on your products. After a fix, the share should shift over the following weeks, not overnight.
  • URL Inspection: test one promoted page and one blocked filter URL. The first should be crawlable with your chosen canonical; the second should say it is blocked by robots.txt.

What would you do?

An outdoor shop’s Monday morning

Illustrative example. A 600-product shop opens Search Console. The Pages report shows 48,000 URLs under “Discovered – currently not indexed”, almost all with ?colour=, ?size= and ?sort=. Crawl stats show requests rising for two months. Three new products added last week have not been crawled yet. Search Console says one filter page, ?feature=waterproof, gets regular clicks.

What should the owner do first?

Before you open Search Console, check your own work in a minute. Paste your promoted page into the free SEO check: it should have a canonical URL, no noindex and no robots.txt block. Then paste one filter address you have just blocked. There, a robots.txt “fail” is the result you want, because it proves the rule catches that parameter.

If Search Console is new to you, the 20-minute tour in the guide on Search Console for business owners shows where each report below lives. And if a page you do want found turns up as blocked or not indexed, the guide Why isn’t my page on Google? walks through the twelve usual causes in order.

Score your shop: the mistakes that quietly cost shops

Each of these is common, and each is easy to undo once you know to look. Answer for your own shop; your result points you at the part of this guide to act on first. “Not sure” counts as no, because finding out is the first fix.

Score yourself

How healthy are your filters?

  1. Filter pages I am removing with noindex stay crawlable until they have left the index

  2. My robots.txt blocks filter parameters by name, not with one broad rule like Disallow: /*?

  3. Page 2 and later of my categories can still be crawled

  4. Filter and sort pages point their canonical at their own category, never the homepage

  5. A filter combination with no products returns a 404, not an empty page with a 200 status

  6. The same selection has only one address, whatever order the filters were ticked in

  7. Every filter page people search for has a clean address, and my filter links point to it

  8. I re-checked robots.txt after my last redesign, theme or platform change

0 / 16

0 of 8 answered. Answer them all to see your result.

Your filter plan for this week

Tick these off as you go. Most shops can do the first five in an afternoon; the rest may need your developer or agency.

Checklist

Filters and faceted navigation

0 of 13 done

Saved on this device only

Questions shop owners ask about filters

Questions people ask

Will blocking filter pages lower my rankings?

Blocking pages that only re-sort or slice the same products should not take anything valuable away. The risk is blocking a filter page that was getting real visits, which is why you check Search Console and promote those first. Nobody can promise a ranking change either way; the aim is that Google spends its time on the pages you want found.

My filters use JavaScript and the address does not change. Am I safe?

If filtering never creates new crawlable links or addresses, there is little for Google to overcrawl. Check by ticking filters and watching the address bar, and look for filter links in your page source. The trade-off is that none of those views can be found in search, so any filter worth ranking still needs its own page.

Should filter pages be in my sitemap?

Only the promoted ones with clean addresses. A sitemap is a list of pages you want indexed, so leave out sort orders and parameter combinations.

How long until I see a difference?

Crawl patterns usually shift over weeks, and the not-indexed counts in the Pages report can take months to settle, because Google revisits low-value URLs rarely. Compare month to month rather than day to day.

While you wait for those numbers to move, there is a payoff you have already set up. Each promoted page is now one address, so every link anyone gives it counts for that one page instead of being split across dozens of copies. Your women’s waterproof boots page is exactly the kind of page a local walking club or a hiking blog might link to.

That is where this series hands over to the next one. The guide on link building that won’t get you in trouble shows how to earn those links without paying for the kind Google ignores or penalises, starting with the links your shop already deserves.

You reached the end

Your summary

This guide has 8 quick questions

Try them to see how much you already know. Your first answer is the one that counts.

Start with the first question
Your checklist0 of 13 done
Revisit the section “Your filter plan for this week”

E-commerce SEO: 0 of 4 read

  1. 1. E-commerce category pages that rank and sell: not read yet
  2. 2. Product pages that get found and bought: not read yet
  3. 3. Out-of-stock and discontinued products: keep, redirect or remove: not read yet
  4. 4. Filters and faceted navigation: keep the useful pages, stop the URL explosion: this guide, not read yet

0 of 57 guides read · Saved on this device only

New guides by email when they are published.

Unsubscribe in one click.

What to do next

  1. Open your biggest category, tick three filters and a sort order, and write down every parameter that appears in the address.
  2. In Search Console, open Indexing, then Pages, and check whether the not-indexed examples are mostly filter URLs.
  3. Pick the one filter people really search for and give it a clean page with its own title, heading and intro before blocking anything.
  4. Paste one promoted page and one blocked filter address into the free SEO check and confirm each says what you decided.

Check your own site

Run our free site check: most of the technical points above, in under a minute.

Run the free check

Links and reputation

Link building that won't get you in trouble

Most “cheap backlinks” emails contain one word that, next to a price, tells you the deal breaks Google’s rules. Can you spot it?

29 min readStart reading

See it applied in an illustrative case study

Invented businesses and modelled figures, not client results: how the work in this guide fits into a whole engagement.

See all 75 studies