How organic clicks grew 39% on a 6,000-product e-shop

The problem

The catalogue existed. Nobody could find it.

The text was the supplier’s, the same text ten other stores had, often with incorrect technical specifications. The category structure had been built with the logic of the warehouse, not the logic of a customer searching. The English side was a translation. And there was nothing structured that an AI tool could retrieve, at a time when more and more people start their shopping there.

What we did

From January 2026 onwards, category by category:

• Restructuring. The categories were rebuilt around the buyer’s logic. Every URL change got a 301 redirect, so nothing was lost from the page history.
• Bilingual content, with a glossary. More than 1,000 pages in Greek and English, with consistent terminology across the whole catalogue. Where needed, local terminology: in Cyprus, solder isn’t called “σύρμα κόλλησης”.
• Specifications verified against the manufacturer. Real errors were found and fixed: a solder alloy that the supplier’s datasheet listed with a melting point of 240 °C when the correct figure is 227 °C, and a set advertised with 30 figures that contained 22.
• Comparison tables between sibling products. Which diameter for which job, which board versus which. It’s what the buyer is looking for, and what AI tools retrieve.
• Structured data. Product, FAQPage and Breadcrumb across the whole catalogue.

The results

The most important number isn’t the growth. It’s the comparison.

We split the 1,493 catalogue pages that appear in Search Console into two groups. The three home pages were excluded, because their traffic was driven by advertising activity rather than by content.

	Pages	Clicks	Impressions
Rewritten categories	911	6,622 → 9,217 · +39%	+42%
Untouched categories	582	3,004 → 2,964 · −1%	−1%

Same store, same period, same seasonality, same ad spend. Forty points of difference.

310 pages that didn’t bring in a single click now bring in 1,691.

At the level of individual pages:

Page	Clicks	CTR
Plastic sheets (EN)	132 → 546	—
Clays & sculpting (EN)	10 → 156	4.3% → 10.3%
Canvases (EN)	2 → 107	3.9% → 11.3%
Mosaics (EN)	28 → 122	7.5% → 12.3%
Foamboard (EN)	8 → 101	3.1% → 9.5%

On the pages that kept the same URL, CTR rose 60%, from 4.72% to 7.55%. Same position in the results, more clicks: the text did the work, not the ranking.

Year on year, from Google Analytics: organic traffic +52%, organic conversions +75%, with less advertising support than the previous year, which included an external ad agency at €1,400 a month.

What we learned

Volume isn’t customers.

A product page for an ESP32 development board became the top organic page of the entire store: 835 visits in three months, from zero. An impressive number for a report.

But we looked at what those visitors did: 21 seconds average time, 0.72% action rate, zero sales. It’s international developer traffic that doesn’t buy from Cyprus.

The opposite example, from the same period: on an educational robotics kit, impressions fell from 1,026 to 468, but the position went from 17 to 8.7, CTR from 1.5% to 5.6%, and clicks went up. Fewer but better-matched impressions.

That’s why our service is split in two: the products that sell are written by hand, and the rest of the catalogue goes through a pipeline. Volume only makes sense as coverage. You don’t pay the price of handwork for it.

Methodology & limitations

This section exists because a case study without it isn’t measurable, it’s advertising.

• It’s our own store. It isn’t an independent client.
• Comparison period: 16/1–16/4/2026 versus 16/6–16/9/2026, Google Search Console. There is no year-on-year comparison because Search Console was connected in January 2026.
• Seasonality works against us. The second window contains July and August, the store’s two weakest months. That’s why the control group of 582 pages exists: it goes through the same seasonality, and it stayed flat.
• The restructuring and the enrichment happened together. We don’t separate their effects, and we don’t claim the result is down to the text alone.
• The numbers are conservative. The craft categories were rewritten in February, that is, inside the baseline period, and electronics were completed inside the measurement period.
• We don’t report revenue. Tracking changed between the two years and channel attribution isn’t comparable. We only publish what can be measured reliably.
• One category, school-equipment, grew 121%, but that is largely down to one product with strong school-season demand. We don’t present it as a typical result.
• In January 2027 Search Console will have a full year of data, and this page will be updated with a proper year-on-year comparison.
• It’s one store. The numbers cover this catalogue and the period we measured. They are not a guarantee for another store. What to expect for yours is something we say after the mini-audit.