Case Studies → Mister Loft
E-CommerceTechnical SEOCanonical ArchitectureContentGEO / AI Search

Mister Loft — Full-Scope SEO
for a Catalog of 297,000 Product Variations

Mister Loft is a Kyiv furniture factory selling loft-style office furniture online. We audited the store, rebuilt the canonical URL architecture, collected the semantic core, cleaned up metadata, improved speed and structured data, wrote category texts and articles, and launched a new “made-to-order furniture” section.

297,513product variations reduced to 935 canonical models
48,882unique search queries collected and sorted
253,692metadata records cleaned and fixed
26expert articles plus a landing page for the new section

THE CHALLENGE

Client’s Challenge

After moving from an old platform, every colour and size of a desk lived on its own URL: hundreds of thousands of near-identical pages competing with each other.

The order had to be restored on a live store, without stopping sales and without losing the pages that already bring buyers.

  • Hundreds of thousands of “colour × size” URLs diluting the catalog
  • Legacy pages and a second language version left by the old platform
  • Metadata with internal codes and duplicated fragments
  • A live store: sales cannot stop
ClientMister Loft
IndustryFurniture manufacturer, e-commerce, Kyiv, Ukraine
ServicesSEO audit, URL architecture, semantic core, metadata, speed, content, landing page
PlatformE-commerce platform as is, schema.org JSON-LD, WebP, Search Console

OUR METHODOLOGY

How We Worked

01

Audit and relevance map

Technical audit and a query-to-page map: 6,807 pairs from Search Console data, a list of duplicates and a decision table.

02

Canonical URL architecture

297,513 “colour × size” variations reduced to 935 representative models; one model — one canonical address.

03

Catalog structure and redirects

250 technical model pages and old landing pages moved with 301 redirects; the site consolidated into one language version (36 legacy pages merged).

04

Semantic core

48,882 queries from several independent sources, grouped and distributed across categories.

05

Metadata and data quality

Title and description of 289 categories rewritten under single rules; 253,692 metadata records fixed; internal codes removed from 311,461 product names.

06

Speed, structured data, AI search

Deferred third-party widgets, WebP images; schema.org JSON-LD: Product with 40 attributes, Organization, BlogPosting, FAQPage.

07

Content and the new section

Texts for 13 key categories, 26 expert articles and a “Made-to-order furniture” landing page with entry points from the catalog and the blog.

PROOF OF WORK

Our Implementation

WHAT WE DELIVERED

Scope of Work

  • Technical audit and relevance map (6,807 pairs)
  • Canonical architecture: 297,513 variations → 935 models
  • 301 redirects for 250 technical and legacy pages
  • One language version instead of two
  • Semantic core of 48,882 queries
  • Metadata for 289 categories and 253,692 fixed records
  • Speed work: deferred scripts, WebP, layout stability
  • schema.org JSON-LD for products, organisation, articles and FAQ
  • Texts for 13 categories and 26 expert articles
  • New “Made-to-order furniture” section with a landing page
  • Internal linking and business-directory profiles