01
The client
Goldia sells fine jewelry it designs and makes itself, with a very large Shopify catalog. Jewelry buyers compare a lot before they spend, so the store needs product and category pages that rank for buyers who are ready to purchase, and a checkout that does not lose them at the last step.
02
The challenge
Goldia first hired us to fix indexing and on page issues they had spotted in Semrush. Once we pulled the real data, the problem was much bigger than a quick fix.
A huge catalog that did not pull traffic. Thousands of Shopify product pages, with many products not ranking or bringing any visitors from search.
URLs that should never have existed. Large numbers of unneeded URLs had been created on the site, wasting crawl budget and diluting the pages that matter.
Cannibalization between products. Similar products competed with each other for the same searches, so none of them ranked well. This was a main reason traffic was lost.
Technical and on page gaps at scale. Weak internal linking, missing or duplicate titles and meta descriptions, and product schema problems across the catalog.
Development and UX issues on product pages. Overlapping UI elements on single product pages and a long buying form made it harder for ready buyers to complete a purchase.
03
What we did
We focused on buyers, not just traffic. Every fix aimed at people ready to purchase, from the technical base up to the checkout.
Technical gap audit
We mapped the real technical gaps first: index bloat, unneeded URLs, crawl waste and broken signals across the catalog.
Buyer intent audience
We found the traffic and audience that is ready to buy, and targeted them instead of broad informational traffic.
URL cleanup and cannibalization fixes
Unneeded URLs cleaned up and competing product pages consolidated so each search has one clear page.
Entity and topical authority
Semantic SEO across categories and products so Google and AI tools understand what Goldia sells and trust it as a source.
Menu structure and on page SEO
A clearer menu and category structure, plus titles, meta descriptions and internal links fixed at scale.
Product schema
Product schema fixed across the catalog so listings can show rich results and AI tools read product details correctly.
Developer guidelines
We guided the developers to fix the development and UI issues on single product pages.
Answer-first blogs
Blog content written answer first, so AI tools can quote it when buyers research jewelry before they buy.
Shorter buying form
We reduced the fields in the buying process so ready buyers finish their order instead of dropping off.
The advanced layer
Beyond basic SEO
Index bloat cleanup
Unneeded URLs removed from the index so crawl budget goes to real products.
Cannibalization mapping
One clear page per search intent across thousands of products.
Buyer intent keyword map
Keywords grouped by readiness to buy, not just volume.
Topical map
Every category, material and style mapped as entities.
Internal link architecture
Guides feed categories, categories feed products.
Product and Organization schema
Structured data that search and AI read correctly.
Answer-first blocks
Short, quotable answers in blogs and buying guides.
Checkout friction audit
Fewer fields, fewer drop offs at the last step.
| Before (Jun 2025) | After (Nov 2026) | |
|---|---|---|
| AI mentions | 0 | 151 |
| Pages cited by AI | 0 | 218 |
| ChatGPT | 0 mentions | 78 mentions |
| Google AI Overviews | 0 mentions | 43 mentions |
| Google AI Mode | 0 mentions | 19 mentions |
| Gemini | 0 mentions | 11 mentions |
| AI Visibility score | 0 | 22 |
| Organic traffic | 721 a month | 1.6K a month |
| Organic keywords | 6.2K | 9.9K |
| Catalog health | Unneeded URLs and product cannibalization | Cleaned up, one clear page per search |
| Checkout | Long buying form | Shorter form, fewer drop offs |
04
The result
Between June 2025 and November 2026, Goldia went from zero AI mentions to 151, with 218 pages cited across ChatGPT, Google AI Overviews, AI Mode and Gemini. Organic traffic grew from 721 to 1.6K a month (+122%) and organic keywords from 6.2K to 9.9K (Semrush, US).
ChatGPT
78
mentions, up from 0
AI Overviews
43
mentions, up from 0
AI Mode
19
mentions, up from 0
Gemini
11
mentions, up from 0
“
I hired Efte for indexing issues and on page SEO problems I had found in Semrush. Other freelancers told me it was easy, but he gave me the actual data: lots of URLs that should never have been created, and cannibalization between our products. That is where we were losing traffic. He was honest that SEO is not an overnight change and takes time. So we brought him on, and he is great with his team.
Our takeaway
With a big ecommerce catalog, more pages is not the answer. Clean URLs, one clear page per search and a checkout that does not fight the buyer move revenue. If your store has thousands of products and flat traffic, our SEO service starts with the same audit, or book a 15 minute call.