// Ecommerce and DTC

Technical SEO for Ecommerce and DTC Catalogues

A catalogue writes its own URLs. Filters multiply them, variants duplicate them, and every discontinued SKU leaves one behind. The crawl problem here is arithmetic, and your product feed is a second copy of the catalogue that fails separately from the site.

What changes in Ecommerce and DTC

Take one collection page with five filters on it. Size, colour, brand, price band, material, six values each. Every combination the template is willing to render is a URL of its own. That is 7,776 URLs from a single collection, before anyone adds a sort order or a page number. Put ascending and descending sort on three fields in front of it and the same collection can produce 46,656. Then multiply by how many collections you have.

That is the ecommerce problem in one paragraph. It is arithmetic, and nobody typed any of it. The template did.

The crawl space a filter bar creates

A crawler spending its day on colour plus material plus sort-descending is not spending it on the forty products you launched this month. So the first piece of work on a catalogue is a decision, facet by facet. Which combinations deserve a crawlable URL, and which are a dead end that should never have been linked.

Some facets earn a page. A size filter on a category where buyers search by size is a real query with demand behind it. The page that answers it should be reachable, canonical to itself, and linked from somewhere a crawler goes. The other combinations earn nothing. They get parameter handling, an unlinked control, or a rule that keeps them out of the crawl entirely. Which of the three you get depends on what your platform will let you change.

That last part decides more than anyone likes. On a hosted store you may not own the robots file, the canonical tag, or the URL shape of a variant. The audit has to say what is fixable in your stack and what has to be worked around, before the fix list is written.

The feed is a second site

Your catalogue exists twice. Once as pages, and once as a product feed going to Merchant Center and to whatever marketplaces and ad platforms you sell through. The feed carries its own title, description, price, availability and identifier per SKU. It is usually built by whoever owns the ERP export, and not by whoever owns the site.

So the two copies drift. The feed says in stock at one price and the page says sold out at another. A free listing then sends a buyer to a page that contradicts the thing that sold them. An audit stopping at the HTML has looked at half your catalogue surface and reported on it as though it were all of it.

We read both and diff them per SKU. Where they disagree, the fix is either in the export or in the page, and saying which one is most of the job.

When a SKU dies

Every product you stop selling leaves a URL behind with links pointing at it, a position in search, and possibly a year of reviews. Deleting it throws all of that away. Redirecting the lot to the homepage is worse. A blanket redirect to an unrelated page gets treated as a soft 404, and you lose the signal anyway.

The decision is per product and it stops being hard once somebody writes it down. A seasonal item returning in nine months keeps its URL and says so on the page. A replaced item redirects to its replacement, one hop. A discontinued item with no successor and nothing pointing at it gets a 410 and drops out cleanly. A discontinued item with real inbound links keeps a page telling a buyer what to look at instead.

Put that policy into the platform once and nobody has to make the call again.

Where this touches your resellers

If you also sell through dealers or marketplaces, the same catalogue data reaches listings you do not control. The price on those listings is a different problem with a different owner. Advertised price is an enforcement question and we run it as a separate service. The two share a source, which is worth knowing, because a bad feed export shows up in both places on the same morning.

What we will not build for a catalogue

We do not generate a page per filter combination to catch long-tail queries. It is the oldest trick in ecommerce search and it sits squarely inside Google’s scaled content abuse policy, which is charged to the whole domain. A catalogue already has more URLs than it can support. The work is spending crawl on the right ones.

This is the Ecommerce and DTC view of Technical SEO. That page covers how the work runs whatever the sector.

Want this scoped against your own numbers?

Book a strategy call