Ecommerce SEO case study: fixing crawl traps and duplicate pages

The store and the problem

The client is an international specialty retailer. The store name stays private, and this page shows only what the Search Console screenshot above supports.

The store had a real catalog and real demand, yet Google sent it few visitors for a business of its size. The owner's first idea was more content and more links. A structured technical review pointed somewhere else: Google was spending its visits on URLs that should never have existed, and getting mixed signals about the ones that should.

That pattern is common on stores built on themes that create URLs automatically. The owner never sees the extra addresses, because customers never type them. Googlebot sees all of them.

What the audit found

Each of these is a known problem on its own, and each is part of what I look at in crawl and indexing work. Google's guides on ecommerce URL structure and faceted navigation describe both the duplicate URL issue and the crawling cost of filter URLs. On a catalog site, the three made each other worse.

What we changed, in order

The work ran in a set order: stop the waste first, then send every signal to one URL, then strengthen the paths to the pages that sell.

1. Step 01

   Merge the collection hierarchy

   Duplicate collection paths were merged into one hierarchy, and routing rules were set so the theme could no longer create parallel versions of a category.

   What this finds. One address per collection, with every internal link pointing to it.

2. Step 02

   Close the filter crawl loops

   Filter combinations were stopped from creating indexable URLs, so crawling went to categories and products instead.

   What this finds. A crawl spent on pages that can rank and sell.

3. Step 03

   Add self referencing canonicals

   Every catalog page received a canonical tag pointing to itself, matching the URL used in links and the sitemap.

   What this finds. Clear, consistent signals about which URL is the real one.

4. Step 04

   Clean up structured data

   BreadcrumbList and Product markup were added across the catalog, so Google could read how products sit inside collections.

   What this finds. A machine readable map of the catalog that matches the visible pages.

5. Step 05

   Rebuild internal links around the best collections

   Links were rebuilt from the strongest collections to the products with the best margins, instead of spreading across every generated URL.

   What this finds. More internal weight on the pages that earn money.

How to read the numbers

Comparing the last 3 months of Google Search Console data with the 3 months before, organic clicks rose +572% organic clicks, last 3 months vs the previous 3 months, view evidence. The table above shows the four figures from the screenshot: clicks, impressions, average click through rate, and average position, each before and after.

Impressions grew fastest. That fits the work: once Google stopped wasting visits on duplicates and filters, more of the real catalog was crawled, indexed, and shown in results. Clicks followed as the merged pages gathered the signals that had been split before.

Average position is an average across every search the store appeared for. When a store starts showing for many more long product searches, some at low positions, the average can hide gains on the main category searches. Read it next to impressions and clicks, not on its own. If your own clicks fell rather than stalled, the checks are different, and they are covered in my guide to a drop in organic traffic.

What this case does not prove

Be careful with any case study, including this one.

The screenshot shows a comparison of two 3 month periods without exact dates, so seasonal demand may have played a part. The store sells internationally, so its market, competitors, and search volumes are different from a store selling in Lahore or Karachi. And the work listed here happened together, so this page cannot say which single change mattered most.

A different site, in a different industry, showed a similar pattern when competing pages were merged, which you can read in the business services case. What this case does show is the order of work that made sense for this store, and that structural fixes can move a store's search traffic without writing new pages first. I cannot promise the same numbers for another store, and nobody honest can.

Checks you can run on your own store

• Compare the number of products and categories in your store admin with the number of indexed pages in Search Console
• Search Console's Page indexing report: look for many URLs with filter or sort parameters
• URL Inspection on a key category page: compare the canonical Google chose with the one the page declares
• Open a category from two different menu paths and check whether the address in the browser changes
• Check that your sitemap, your menu links, and your canonical tags all use the same URL for each category

If several of these show problems, your store may have the same kind of structural issue, and ecommerce SEO done end to end is how I handle it. My guide to an online store that does not rank walks through each cause in order, with the check for each one.

What a store owner in Pakistan can take from this

Most stores I see in Pakistan run on Shopify, WooCommerce, or a custom build from a local agency. The same three problems can show up on all of them, because themes and plugins are added without anyone checking the URLs they create.

The lesson is about order. Before you pay for new product descriptions or blog posts, make sure Google sees one clean address for each category and product. New content on a messy structure lands in the same queue of duplicates and filters, and it struggles for the same reasons.

Content still matters once the structure is clean. Category pages that explain the range, and product pages written in your own words, are what turn indexed pages into ranking pages. That is where writing for category and product pages comes in, after the fixes in this case.

If you would rather talk it through before committing to anything, a single paid consultation is enough to read your Search Console data together. For ongoing work, the monthly plans in rupees show what each level includes. The budget calculator for SEO work gives a rough range first. Store owners in Lahore can also see how I work with ecommerce businesses in Lahore.

Your next step

If your store has many more URLs in Google than products in your admin, send me your store link and I will tell you what I see first. I reply within 24 hours. You can also run the free website scan to spot the obvious problems.