Taking a publisher's online bookstore to 3x sales in year one, with organic traffic up 45% in under six months
The situation
A book publisher selling its own catalogue through its own store, in a market where general retailers list the same titles with more authority. Organic search already drove almost 63% of conversions, yet the category pages, the pages that should carry a bookshop, brought in only 7% of organic visits.
The audit explained why. There were no product categories or subcategories at all: filters changed the listing but never produced a page that could rank. 5,696 pages had no meta description and no H1, 4,273 titles were duplicates, product names were not written for search and descriptions barely existed. There was no robots.txt and no XML sitemap, and 57 pages, most of them categories, were being kept out of the index by 302 redirects nobody could explain. Googlebot's view of the site was scattered, with link equity pooling in the wrong places.
What I did
Cannibalisation audit first, then a full site structure with monthly search volume for every category, subcategory and sub-subcategory, and keyword research for every page type: home, categories, filters, products, authors and blog. Filters with real demand became indexable subcategories with their own URLs; the rest stayed out of the index by rule.
On-page structure per page type followed from that research: title, H1 and meta description patterns that generate automatically when a filter page is created, H2 and H3 patterns for the homepage and categories, 250 to 500 words of WDF/IDF-checked copy split above and below each listing, product names moved into the H1, and a new Google-friendly URL structure with a 301 map to be staged on a test server before going live.
Crawl budget and indexation: robots.txt written from scratch, XML sitemaps built and submitted, pagination given its own URLs with noindex/follow and rel prev/next, canonicals on categories and parameter URLs, sorting parameters set to noindex/nofollow, and index/follow rules for every level from category down to product. Structured data per page type: Organization, WebSite and BreadcrumbList sitewide, Product, Offer and AggregateRating on books, and aggregated ratings on categories.
Internal linking and navigation last: the main menu rebuilt around the category tree, low-value pages moved to the footer, and a pre-footer of low-hanging-fruit terms at positions 10 to 20 linking to the exact pages that needed the push. Everything went to the client's team as a prioritised list, with the fastest-moving items first.
What happened
Organic traffic was up 45% inside six months. Sales, the number the client actually tracked and the only KPI the plan was measured on, reached three times the previous level in the first year of the engagement.
What this proved
- On a catalogue site the technical layer decides which pages can rank at all, and the content layer decides whether they win the click. Categories that never existed cannot rank, whatever the copy.
- Pattern-generated titles, H1s and meta descriptions fixed thousands of pages at once, which is the only way a small team ships on-page work at catalogue scale.
- The 45% traffic gain and the 3x sales figure are client-reported and appear here as claims, not as counters.
Client name withheld under NDA. Search Console documentation available on request.
SEO & AEO Consultant
Contact Marko Miljkovic
Send the domain and a sentence on what has stalled. You get a straight answer on what I would check first, and whether I am the right fit, before any call.