I have two thousand products and only a few hundred are indexed. Why?

Usually because the crawler's effort is being consumed by filter and sort URLs that generate near-identical pages by the thousand, leaving the real products under-visited. Sometimes it is thin product descriptions copied from the manufacturer. The first job is establishing which of those it is, because the fixes are completely different.

Should my filters be indexed or not?

Selectively. Combinations people genuinely search for — typically a brand within a category — are worth having as indexable pages. Combinations of colour, size and price are not, and left open they multiply into tens of thousands of near-duplicates. Everything outside the chosen set gets handled with canonical tags or crawl rules.

What happens to products that go out of stock?

It depends on whether they are coming back. Seasonal or temporarily unavailable items keep their page with the status made clear, because that page may hold years of accumulated value. Permanently discontinued items get redirected to the closest relevant product. Deleting the page is almost always the wrong answer.

Does the platform we are on matter?

Yes, and it decides which fixes are available to you at all. Shopify generates its own sitemap and gives you limited control over parts of the URL structure, so certain fixes have to be made in theme files. WooCommerce and Magento give more control and correspondingly more ways to get it wrong. Whether we can edit theme files directly affects both scope and fee.

Can you add review stars to our search results?

We can add valid product markup so the store is eligible for them, and eligibility is genuinely all anyone can offer. What we will not do is output rating markup where there are no real reviews behind it. Fabricated review markup is treated as spam and puts the whole store's rich result eligibility at risk.

Why is this monthly rather than a project?

Because the catalogue changes. New products arrive without descriptions, lines get discontinued, categories get restructured and seasonal ranges come and go. A store that is fixed once drifts back within months, so the work is a continuing process with monthly reporting.

SEO & Organic Growth

E-commerce SEO

Category and product pages fixed at catalogue scale — indexation, faceted navigation, product markup and the out-of-stock decisions that quietly cost stores their rankings.

From: $500–$3,500/mo
Typical time: ongoing
Usually bought by: Online stores

Two thousand products, three hundred indexed

Store owners describe this problem in money, which is one of the better things about working on e-commerce. The conversation is not about traffic in the abstract — it is that category pages which should be earning are sitting on the second page, or that most of the catalogue is not showing up in search at all, or that a replatform happened and something went badly wrong with the old URLs.

The underlying cause is nearly always scale. Everything that is a minor annoyance on a twenty-page site becomes a structural problem across a catalogue. A duplicate title is trivial; four thousand duplicate titles generated by a template is a different thing. A filter that creates a new URL is a convenience; a filter system that can generate tens of thousands of URL combinations consumes the crawling effort that should have been spent on your actual products.

That is the real currency here. A search engine allocates a finite amount of attention to any site, and on a large store that allocation is the constraint. Most e-commerce work is about spending it on the pages that sell rather than on machine-generated variations of them, and then making sure those pages have something on them worth ranking.

How the work runs

Crawl carefully, and establish the indexation picture. Crawling a store without constraints can run indefinitely because the filters keep producing new URLs, so the crawl is configured before it starts. Products are also crawled from an exported list so the real catalogue can be compared against what is actually indexed. The output is a simple, uncomfortable number — how many products exist, and how many are in the index. One to two days.
Decide the faceted navigation rules and apply them. Which filter combinations deserve to be indexable pages, and which must be closed off. Typically brand within a category earns a page because people search that way; colour with size with price does not. The rest is handled with canonical tags and crawl directives rather than left to chance. Two to four days.
Give category pages something to rank with. On most stores the category page is an empty grid of products with a heading, which gives a search engine almost nothing to work from even though these are frequently the highest-value pages on the site. They get genuine introductory content, sensible internal links and a heading structure that reflects how buyers actually narrow down. Ongoing across the priority categories.
Implement product structured data across the catalogue. Generated from your product data rather than typed by hand, including price, currency, availability, brand and image. Aggregate ratings are only ever included where a real rating value and a real review count both exist, empty identifier fields are omitted rather than written blank, and a sample is validated against Google's Rich Results Test before the rest is deployed. Two to five days, depending on catalogue size and platform.
Set the stock and internal linking rules, then run it monthly. A documented decision for discontinued versus temporarily unavailable products, related-product and frequently-bought-together linking, image weight reduced on product templates, and a monthly review as the catalogue changes. Ongoing.

What moves the price

The range is $500–$3,500/mo and the engagement is ongoing, because a catalogue never stops changing. Three things set where a store sits in that range.

How large the catalogue is. A few hundred products is a manageable set. Tens of thousands changes the method entirely — everything has to be generated, batched and sampled rather than handled page by page, and the quality control on that is where the time goes.

The platform, and how deep we can go into it. Some fixes on a hosted platform can only be made inside theme files. Where we have that access we implement directly; where we can only recommend, the work becomes a specification and the timeline depends on your developers.

Whether implementation is included. Producing the markup, the facet rules and the redirect logic is one thing. Deploying them across a live store, verifying them and monitoring for regressions is another, and it is the part that determines whether any of it takes effect.

What usually goes wrong

Every filter combination gets opened up for indexing. It follows from a reasonable-sounding belief that more pages means more traffic. What it produces is tens of thousands of pages that are almost identical to one another, crawling effort spent on noise, and the genuine category pages weakened by the duplication surrounding them. The correct move is counter-intuitive — fewer indexable pages, with the important ones made stronger.

Out-of-stock product pages get deleted. It feels like housekeeping. That page may have accumulated years of links and search history, and deleting it discards all of it and produces an error for anyone who arrives. Redirect it to the nearest equivalent, or keep it and show what to buy instead — but do not simply remove it.

Rating markup is added where there are no ratings. Stars in the search results are tempting enough that stores output rating markup with placeholder or zero values, or apply a site-wide rating to every product. Fabricated review markup is treated as spam, and the consequence is not that the stars fail to appear — it is that the store's eligibility for rich results is put at risk across the whole catalogue.

See where the site stands before you brief anyone.

If the site is already live, start from what it scores rather than from a guess. The audit reports what it checked, what it found, and — the part most tools leave out — what it could not check, and why.

See the website audit Talk about this work

Questions people ask before they commit.

I have two thousand products and only a few hundred are indexed. Why?: Usually because the crawler's effort is being consumed by filter and sort URLs that generate near-identical pages by the thousand, leaving the real products under-visited. Sometimes it is thin product descriptions copied from the manufacturer. The first job is establishing which of those it is, because the fixes are completely different.
Should my filters be indexed or not?: Selectively. Combinations people genuinely search for — typically a brand within a category — are worth having as indexable pages. Combinations of colour, size and price are not, and left open they multiply into tens of thousands of near-duplicates. Everything outside the chosen set gets handled with canonical tags or crawl rules.
What happens to products that go out of stock?: It depends on whether they are coming back. Seasonal or temporarily unavailable items keep their page with the status made clear, because that page may hold years of accumulated value. Permanently discontinued items get redirected to the closest relevant product. Deleting the page is almost always the wrong answer.
Does the platform we are on matter?: Yes, and it decides which fixes are available to you at all. Shopify generates its own sitemap and gives you limited control over parts of the URL structure, so certain fixes have to be made in theme files. WooCommerce and Magento give more control and correspondingly more ways to get it wrong. Whether we can edit theme files directly affects both scope and fee.
Can you add review stars to our search results?: We can add valid product markup so the store is eligible for them, and eligibility is genuinely all anyone can offer. What we will not do is output rating markup where there are no real reviews behind it. Fabricated review markup is treated as spam and puts the whole store's rich result eligibility at risk.
Why is this monthly rather than a project?: Because the catalogue changes. New products arrive without descriptions, lines get discontinued, categories get restructured and seasonal ranges come and go. A store that is fixed once drifts back within months, so the work is a continuing process with monthly reporting.

Often taken together.

Start with the brief, not the invoice.

Send what you are trying to achieve. You get a written scope, a price and a date — or an honest “this is not the right service for that”.

Talk about E-commerce SEO All SEO & Organic Growth services