How to Fix Crawl Budget Waste from URL Parameters

Crawl budget waste from parameters happens when a large share of your internal links point to URLs with query strings, such as ?sort=price, ?color=red or ?utm_source=nav. Each one looks like a separate page to a crawler. By the end of this guide, fewer than 10% of your internal links will point to parameterised URLs, and the parameter URLs that remain will be blocked or canonicalised.

By Updated ⏱ 1-3 hours (longer on large stores with faceted navigation)🎯 Intermediate● Medium impactTechnical
Illustration: Crawl budget waste (parameters) fix guide
TECHNICAL · FIX GUIDE Crawl budget waste (parameters) IMPACTEFFORT TIME1-3 hours (longer on large stores with faceted navigation) LEVELIntermediate

Why it matters

sessionid= variant is a separate URL that Googlebot and AI crawlers may fetch. On sites with faceted navigation, these combinations multiply into thousands of near-duplicate pages. Crawlers then spend their visits on those duplicates instead of new products or articles, so fresh content gets indexed later.

Ranking signals also get split across duplicate URLs. Internal links with UTM tags cause a further problem: they overwrite the original traffic source in GA4, so your attribution data becomes wrong.

✕Before38% of internal links point toparameterised URLs, e.g./shoes?sort=price_asc&color=red and✓After6% of internal links point toparameterised URLs; sort and filtercontrols are buttons, and the

What you need

Access to your theme templates or front-end code
Ability to edit robots.txt (or a platform setting that controls it)
A crawler such as Screaming Frog SEO Spider or Sitebulb, or a browser console
Google Search Console access for the property

Step-by-step fix

1

Measure which internal links carry parameters

and divides that by all internal links. There is one exception: plain pagination is ignored when the URL ends in page=` followed by a number. Anything above 10% is flagged, and above 30% fails.

`. For a quick check of a single template, paste the snippet below into the browser console on a category page, a product page and a blog post.

javascript
const links=[...document.querySelectorAll('a[href]')].filter(a=>a.host===location.host);
const q=links.filter(a=>a.href.includes('?')&&!/[?&]page=\d+$/.test(a.href));
console.log(Math.round(100*q.length/Math.max(1,links.length))+'% parameterised');
console.table(q.map(a=>a.href));
Pro tip: Sort the export by parameter name rather than by URL. A single template, such as a product card or a footer, is usually responsible for most of the count.
2

Group the parameters by type

List every parameter you found and decide what to do with each type:

  • Tracking (utm_*, ref, source, gclid on internal links): remove them.
  • Session/state (sessionid, PHPSESSID, currency, cart): move them to cookies.
  • Sort/view (sort, orderby, view, limit): turn the links into buttons.
  • Filters/facets (color, size, filter.*): use clean paths for facets worth indexing and buttons for the rest.
  • Legacy IDs (?p=123, ?id=45, ?variant=): replace them with the canonical clean URL.
Pro tip: Pagination such as `?page=2` only passes the check when `page=N` is at the very end of the URL. `/shoes?page=2&sort=new` is still counted, and so is `/shoes?page=2#top`.
3

Remove tracking and session parameters from links

Search your templates, menus and content for internal links containing utm_, ref= or session IDs, and replace them with the bare URL. If you need to know which banner or menu item drove a click, track it as a GA4 event or with a data- attribute instead. Turn off URL-based session handling on the server, for example by setting session.use_trans_sid = 0 in PHP.

html
<!-- Before -->
<a href="/sale?utm_source=homepage&utm_medium=banner">Sale</a>
<!-- After -->
<a href="/sale/" data-track="home-banner-sale">Sale</a>
Pro tip: You can bulk-find these in a WordPress database with `SELECT ID FROM wp_posts WHERE post_content LIKE '%utm_source%';`.
4

Turn sort and filter links into buttons or forms

Sort orders and most filter combinations should not be crawlable links. Render them as a <select>, as <button> elements, or as a GET <form>. All three still update the URL for users and keep it shareable, but they are not <a href> links, so crawlers do not follow them.

For facets that have real search demand, such as a colour plus a category, link to a clean static path like /shoes/red/ and give that page its own title and H1.

html
<form method="get" action="/shoes/">
  <label for="sort">Sort by</label>
  <select id="sort" name="sort" onchange="this.form.submit()">
    <option value="">Featured</option>
    <option value="price_asc">Price: low to high</option>
  </select>
  <noscript><button type="submit">Apply</button></noscript>
</form>
Pro tip: Do not hide filter links with JavaScript-generated `href="#"` tricks while keeping the real URL in the DOM. Build the URL when the user clicks.
5

Point cards and menus at clean URLs

p=45. Edit those templates so they output the canonical permalink. p=ID.

If you paginate with parameters, make sure page=N is the last parameter, or switch to a path format like /blog/page/2/.

6

Block or canonicalise the parameter URLs that remain

Some parameter URLs will still exist, because users share them and old links point to them. Handle each pattern one of two ways:

  • Pure crawl waste with no backlinks (sort orders, multi-filter combinations, add-to-cart URLs): Disallow them in robots.txt.
  • Patterns with external links or shared traffic: leave them crawlable and add <link rel="canonical"> pointing to the clean URL.

Do not use both methods on the same pattern. If robots.txt blocks a URL, Google cannot fetch the page, so it never sees the canonical tag. Google retired the Search Console URL Parameters tool in 2022, so robots.txt and canonicals are the controls you have now.

text
User-agent: *
Disallow: /*?*sort=
Disallow: /*?*orderby=
Disallow: /*?*add-to-cart=
Disallow: /*?*sessionid=
Disallow: /*?*filter_*&filter_

Sitemap: https://example.com/sitemap.xml
Pro tip: robots.txt only stops crawling. It does not change your internal-link percentage, so steps 3-5 are what actually move the score.
7

Recrawl and confirm the share is under 10%

Clear any page and CDN caches, then run the same crawl again. in the Destination column (excluding URLs that end in page=N`) by the total number of internal-link rows. Also run the console snippet again on your main templates.

Finally, rerun the audit and confirm the result is 10% or lower.

Pro tip: In Search Console, check **Settings > Crawl stats > By file type / By purpose** over the next 2-4 weeks. The share of requests going to parameter URLs should drop.

Platform-specific notes

WordPress / WooCommerce

Shop archives output add-to-cart buttons as `<a href="?add-to-cart=ID">`, and layered-nav widgets output `?filter_color=` links. Both count as parameterised links. Either replace the archive button template (`loop/add-to-cart.php` in your child theme) with a button, or switch to block-based Product Filters that use buttons. Disallow `add-to-cart` in robots.txt. Yoast and Rank Math already canonicalise `?orderby=` to the base archive.

Shopify

Shopify's default robots.txt already disallows most `sort_by` and `filter` combinations. To customise it, go to **Online Store > Themes > Edit code** and add `robots.txt.liquid`. The bigger source of parameterised links is usually product cards linking to `{{ product.url }}?variant=`. Edit the card snippet (often `card-product.liquid`) to use `{{ product.url }}` only.

Next.js

Avoid `<Link href={{ pathname: '/shop', query: { sort } }}>` for sort and filter controls. Use a `<select>` or `<button>` that calls `router.push()` or `router.replace()` from `next/navigation` instead. Model indexable facets as dynamic route segments (`app/shop/[category]/[color]/page.tsx`), and set `alternates.canonical` in `generateMetadata`.

Webflow

CMS filter libraries such as Finsweet Attributes write state to query strings. Keep filters as buttons or checkboxes rather than link blocks. Edit robots.txt under **Site settings > SEO > Indexing**, and set per-page canonicals in the page settings.

Nginx

To clean up old tracking links, you can 301 internal requests that carry `utm_` parameters to the bare path, using a `map` on `$args` plus `return 301 $uri;`. Do this only on internal paths, and only if you no longer need those parameters for campaign attribution. A redirect does not remove the links themselves, so you still need to fix the templates.

Check that it worked

✅ How to verify

  • Rerun the audit and confirm it reports 10% or less of internal links pointing to parameterised URLs.
  • Run the console snippet on a category, product and blog page and confirm the parameterised share is under 10% on each.
  • Use Search Console's URL Inspection tool on a `?sort=` URL and confirm it shows as blocked by robots.txt, or as an alternate page with the correct canonical.
  • After 2-4 weeks, check **Settings > Crawl stats** in Search Console for fewer crawl requests to parameter URLs.

⚠️ Common mistakes

  • Adding robots.txt rules but leaving the parameterised `<a href>` links in templates, so the internal-link share does not change.
  • Blocking a URL pattern in robots.txt and also relying on its canonical tag, which Google never sees because it cannot crawl the page.
  • Putting `page=N` before other parameters (e.g. `?page=2&sort=new`), so pagination links are still counted as parameter waste.
  • Adding UTM tags to internal banners and menus, which wastes crawl and overwrites the original traffic source in GA4.

Is this still an issue on your site? Run a free audit — ZORO checks this and 294 other SEO, AI-visibility, accessibility and security signals, with fixes for your platform.

Check my site free

FAQ

Should I block URL parameters in robots.txt or use canonical tags?

Use robots.txt for parameter patterns that are pure crawl waste and have no backlinks, such as sort orders and add-to-cart URLs. Use canonicals for parameter URLs that users share or that other sites link to. Do not apply both to the same pattern.

Do UTM parameters on internal links hurt SEO?

Yes. They create duplicate URLs for crawlers, and they reset the session's source in analytics. Track internal clicks with events instead.

Is ?page=2 pagination a crawl budget problem?

Simple pagination is fine, and this check ignores it when the URL ends in `page=` and a number. Problems start when pagination is combined with sort or filter parameters.

How do I stop Google crawling faceted navigation?

Render filter and sort controls as buttons or forms rather than links. Give only the facets with real search demand clean static URLs. Disallow the remaining multi-filter patterns in robots.txt.