Skip to content
Get Ranked On TopIllustrative case studyCommerce SEO plan

A fictional business. The figures are a modelled scenario, not a client result.

Hemlock & Hinge

8,000-SKU cabinet hardware store · Seattle

Ecommerce case notes · Seattle · North America

How a cabinet hardware store recut its catalogue to fit the way people search

+64% organic revenue once every search had its own page

12 months · North America · Commerce SEO plan

  • Keyword-led taxonomy
  • Faceted navigation rules
  • Product & variant schema
  • Category copy & internal links
/matte-black//128mm//3-inch//soft-close//brass-128mm/matte black128mm3 inchsoft closebrass 128mmTAILS · WHAT PEOPLE SEARCHPINS · THE CATEGORY PAGESCut to fit: one search, one pagePlate I. The catalogue, recut so every search seats in its own page.
Organic revenue (month 12)+64%

01The challenge

Cut by collection. Searched by finish and size.

Categories were named after in-house collections, not the finishes and centre-to-centre sizes shoppers type, so 8,000 SKUs ranked for nothing specific.

Hemlock & Hinge sells 8,000 pulls, knobs and hinges to renovators and cabinet makers across the US. Its catalogue was organised the way the warehouse thought: by in-house collections and part numbers. Shoppers think in finishes and hole spacing. They measure the old pull, type “128mm” or “3 inch”, add “matte black”, and buy from whoever has a page for exactly that. Hemlock & Hinge had the products, but no page that said so.

Fig. 2 · One pull, two vocabularies

128 mm centre to centreHH-4410-MB
The catalogue said
HH-4410-MB · Ballard Series
The shopper typed
matte black cabinet pull 128mm

No page on the site was cut for that search, so Google sent it to competitors who had a matte-black page and a 128 mm page.

71%of 41,000 mapped queries name a finish or a centre-to-centre size. Not one category page targeted either.

02Diagnosis

What the audit found on the bench.

Taxonomy

26 of 36

26 of the 36 category pages were named after in-house collections (Ballard, Alki, Fremont) that nobody searches for; 19 earned no impressions in a year. The other ten were bare product types: Pulls, Knobs, Hinges.

Entry 1 of 4

Faceted navigation

185,964

Seven filters (finish, size, material, style, collection, price, sort) combined freely into 185,964 crawlable parameter URLs, all with the same title. Googlebot spent 78% of its requests there.

Entry 2 of 4

Search demand

71%

41,000 queries mapped to the catalogue: 71% named a finish or a centre-to-centre size. Hinge searches named a mechanism (soft-close, concealed, overlay) the site never used as a page.

Entry 3 of 4

Product pages

8,000 SKUs

Titles were part numbers. Finish and size lived only in a variant dropdown, so Google saw one page per pull, no Product schema, and 1,400 near-duplicates that differed only by finish.

Entry 4 of 4

The cut didn’t match the demand

What shoppers search for · 41,000 queries
How the 36 category pages were cut, before
How the 200 category pages are cut, month 12
  • Finish
  • Centre-to-centre size
  • Finish + size
  • Hinge type
  • In-house collection
  • Product type only

03The strategy

Four cuts, one joint.

Built on our Commerce SEO plan: ecommerce SEO for catalogues where the category tree, the filters and the product data decide what can rank.

  1. Cut 1

    Keyword-led taxonomy

    A new tree cut the way people search: product type first, then finish, centre-to-centre size and hinge type. Finish × size pages exist only where search volume proved demand: 90 of 612 possible. The 26 collection pages were 301-redirected to the finish pages that now hold their products.

  2. Cut 2

    Faceted navigation rules

    Filters stayed for shoppers but left the index. Parameter URLs carry a canonical to their category, sort and price are noindex, and only the 190 combinations with real demand became static, linkable paths. 185,964 crawlable filter URLs fell to none.

  3. Cut 3

    Product & variant schema

    One ProductGroup per pull, every finish and size a variant with its own GTIN, price and stock, declared through variesBy. Titles are now built from attributes, such as Matte Black Bar Pull, 128 mm (5 in), so each variant can match the search that names it.

  4. Cut 4

    Category copy & internal links

    Each category page got 120–200 words that answer the real question (which spacing suits a 24-inch drawer, which finishes match), plus a size guide, a finish comparison and links from every product back up to its finish and size pages.

04The dovetail jig

Recut the category tree until every search locks in.

Each tail is a cluster of real searches. Each pin is a kind of category page. Switch the cuts on and off: a tail with no socket holds the boards apart, and a cut that serves no search leaves a gap in the joint.

12345678TAILS · SEARCHESPINS · CATEGORY PAGES

Keyword-led cuts

Legacy cuts

8 searches have no socket, so the boards won’t close.

Demand with its own page
0%
Category pages
36
Indexable URLs
194,000

The eight tails

  1. matte black cabinet pulls14,800 / mo

    Falls back to /collections/ballard/

  2. 3 inch drawer pulls9,900 / mo

    Falls back to /pulls/?size=76

  3. 128mm cabinet pulls6,600 / mo

    Falls back to /pulls/?size=128&sort=new

  4. brushed brass pulls 128mm2,400 / mo

    Falls back to /collections/alki/

  5. oil rubbed bronze cabinet knobs5,400 / mo

    Falls back to /collections/fremont/

  6. soft close cabinet hinges12,100 / mo

    Falls back to /hinges/

  7. concealed hinges for frameless cabinets4,400 / mo

    Falls back to /hinges/?type=euro&page=3

  8. satin nickel cup pulls2,900 / mo

    Falls back to /collections/ballard/?page=2

Volumes are modelled US monthly searches. Collections and filters are the old store’s cuts; switch them off and the finish, size and hinge pages take their place.

05Month by month

12 months on the bench.

  1. Months 1–2

    Measure

    41,000 queries mapped to 8,000 SKUs, a crawl of 194,000 URLs, twelve weeks of server logs and revenue per category from order data.

  2. Months 3–4

    Mark out

    The new tree drawn on paper first: type, finish, size, hinge type. 90 finish × size pages chosen from 612 on search volume; a 301 map for 26 collections.

  3. Month 5

    Clear the waste

    Canonical and noindex rules for filters, sort and price. Googlebot’s share on parameter URLs fell from 74% to 41% in four weeks.

  4. Months 6–8

    Cut and fit

    200 category pages launched, ProductGroup schema on 8,000 SKUs, titles rebuilt from attributes in batches, one product type at a time.

  5. Months 9–12

    Glue up

    Category copy, size guides and finish comparisons; links from products up to their finish and size pages; merchant listings checked monthly.

Googlebot requests spent on filter URLs78% → 8%

06Results

A joint that closes, and stays closed.

Organic revenue (month 12)

+64%

1,180 → 1,935 organic orders / month

Non-brand organic clicks / month

21,400 → 46,900

+119% in twelve months

Category pages in the top 10

3 → 146

146 of 200 new category pages

Indexable URLs

194,000 → 8,200

Filters out of the index; crawl back on products and categories

Modelled scenario for a fictional business — not a client result. Rankings and traffic are never guaranteed; see our Terms.

Organic revenue and non-brand clicksRevenue index (month 1 = 100)Non-brand clicks / month
NEW TREE LIVE100140180025K50KM1M2M3M4M5M6M7M8M9M10M11M12164

The eight tails, ranked

SearchCategory pageVolume / moBeforeMonth 12
matte black cabinet pulls/pulls/matte-black/14,800#47#4
3 inch drawer pulls/pulls/3-inch-76mm/9,900100+#6
128mm cabinet pulls/pulls/128mm-5-inch/6,600#62#3
brushed brass pulls 128mm/pulls/brushed-brass/128mm/2,400100+#2
oil rubbed bronze cabinet knobs/knobs/oil-rubbed-bronze/5,400#38#5
soft close cabinet hinges/hinges/soft-close/12,100#29#7
concealed hinges for frameless cabinets/hinges/concealed-frameless/4,400100+#3
satin nickel cup pulls/cup-pulls/satin-nickel/2,900#54#2

Modelled Google US positions and monthly volumes; 100+ means outside the first hundred results. Queries shown as shoppers type them.

Get a plan like this

Running a catalogue that ranks for nothing specific? Start with the Commerce SEO plan, or tell us how your shoppers search.

← All case studies