Illustrative case study

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

A case study in

Fallowmere Press

Independent literary publisher · Edinburgh

How a small Scottish press stopped losing its own readers to the big online bookstores

+178%

direct book sales, over 12 months

Get Ranked On Top · Edinburgh · UK & Ireland

Colophon
Client — Fallowmere Press
Sector — Independent literary publisher · Edinburgh
Industry — Creative & Media
Region — Edinburgh, United Kingdom
Plan — Content Engine
Engagement — 12 months
Set in — the reader’s own system serif

“A book that cannot be found has, in every way that matters to its reader, not yet been published.”— from the brief

Chapter I

The Challenge

Searches for Fallowmere's authors and titles led straight to retailers; the press's site had no author pages, reading guides or book structured data.

Fallowmere publishes around twenty titles a year from a converted printworks in Leith: literary fiction, poetry and translated crime. Its readers were loyal, but they rarely arrived at its door. Someone hearing an author on the radio searched the name, and the first five results belonged to retailers that paid the press a fraction of the cover price and kept the reader’s email address.

The aim was not to beat the large bookstores at selling books. It was to make the press the authoritative source for its own authors and titles, so that search engines and AI assistants would point to it first, and to make buying direct the easy, obvious choice once a reader arrived.

Chapter II

A Diagnosis

A four-week audit covered crawl logs, structured data, the catalogue and 120 reader-style prompts in AI assistants. Six findings shaped the plan.

  1. §1 Entity

    No author pages at all — 31 living authors existed only as a name in a book's blurb.

    Search engines had nothing on fallowmerepress.co.uk to reconcile an author entity against, so panels were built from retailer and encyclopaedia data.

  2. §2 Structured data

    Zero Book, Person or Organization markup across 214 title pages.

    No ISBN, format, author or offer properties: retailers' pages were the only machine-readable record of each edition.

  3. §3 Content

    Title pages carried 60–90 words of jacket copy and a buy button pointing off-site.

    Pages answered no reader question — who is this for, what to read next, discussion points for book clubs.

  4. §4 Technical

    Backlist hidden behind a JavaScript-only filter; 138 of 214 titles not indexed.

    Crawl logs showed bots reaching the catalogue root and stopping.

  5. §5 Commerce

    Direct checkout existed but was linked from only 9% of title pages.

    Every 'Buy' route defaulted to a large retailer, giving away margin and the reader relationship.

  6. §6 AI search

    AI answers for 'books like…' queries cited retailers and review aggregators, never the press.

    No quotable, well-structured passages existed for an assistant to cite.

The strategy runs to four chapters, one per tactic, built on our Content Engine plan (content & ai search).

Chapter III

Author entity pages

A permanent home for every author: biography, full bibliography, interviews and events, with consistent naming and sameAs links to the authority records search engines already trust.1

  1. a.31 author pages, each with a canonical name form, portrait alt text and dated biography
  2. b.Person schema with sameAs to library authority files and the author's own site
  3. c.Bibliography blocks linking every edition, so author and books reinforce each other

Plate 1 — the same search, a year apart

the salt orchard isla marrick
  1. big-online-bookstore.example

    The Salt Orchard: Paperback — free delivery

  2. chain-bookshop.example

    The Salt Orchard by Isla Marrick | Buy online

  3. reader-reviews.example

    The Salt Orchard — reviews and ratings

  4. secondhand-books.example

    Salt Orchard Marrick — used from £2.10

No knowledge panel. The author’s details exist only on retailer listings.

Mock-up of a results page for a fictional title. Other sites are placeholders.

Chapter IV

Book schema

Every title page rebuilt as the definitive record of the work: each edition with ISBN, format, page count, publication date and a direct offer, marked up as Book with workExample editions.2

  1. a.Book + workExample markup for 214 titles and 352 editions, validated in CI
  2. b.Direct-checkout Offer on every edition; retailer links kept but moved below
  3. c.Static, crawlable backlist pages replacing the JavaScript filter

Chapter V

Topic clusters

Reading guides grouped into clusters around what readers actually search — Scottish landscape fiction, translated Nordic crime, debut poetry — each linking guides, titles and authors.3

  1. a.Six cluster hubs, each owning a reader intent rather than a list of books
  2. b.40 reading guides with discussion questions, context essays and 'read next' paths
  3. c.Internal links from every guide to the title page with direct checkout

Chapter VI

AI Overviews

Passages written to be cited: short, factual answers to 'books like…', 'what to read after…' and book-club questions, placed high on guide pages with clear attribution.4

  1. a.Answer-first paragraphs of 40–60 words on every guide and cluster hub
  2. b.FAQ blocks drawn from real book-club enquiries the press had received
  3. c.Monthly tracking of 120 prompts across AI answers and overviews

Chapter VII

The 12 Months

  1. ICrawl, log and schema audit; entity map of 31 authors; baseline of 120 AI prompts.0
  2. IIStatic backlist replaces JS filter; canonical author names agreed with the editors.2
  3. IIIFirst 12 author pages live with Person schema; first two reading guides.5
  4. IVBook schema on all 214 titles; direct-checkout offers added to every edition.9
  5. VRemaining author pages; first cluster hub, Scottish landscape fiction.13
  6. VITwo further cluster hubs; answer-first rewrite of top 30 title pages.17
  7. VIIBook-club FAQ blocks; events calendar marked up and linked to authors.21
  8. VIIITranslated-fiction cluster; spring list launched with full pages before publication.25
  9. IXInternal-link pass across guides; checkout simplified to two steps.29
  10. XPoetry and essays clusters; AI prompt set expanded to holiday-gifting queries.33
  11. XIAutumn list launch pages; guide refresh based on search console queries.37
  12. XIIFinal guides published; year-end review and the next year's plan.40

Bars: reading guides published to date.

Chapter VIII

Results

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

+178%
Direct book sales
2 → 19
Author knowledge panels
40
Reading guides published
+231%
Organic sessions
Organic sessions, monthly — +231%
04k8k12k012345678910111210,261
Direct book sales, orders a month — +178%
01002003004000123456789101112389
Author knowledge panels: 2 → 19 — one spine per author

Cream: panels at the start · Rose: gained during the engagement · Dark: not yet

Reading guides published: 40 — leaves bound by month
Rankings for the press’s own pages, month 0 to month 12
QueryIntentBeforeAfter
isla marrickAuthor91
the salt orchard novelTitle61
the salt orchard reading guideBook club—1
scottish landscape fictionTopic—3
books like the salt orchardAI answer—2
independent publisher edinburghBrand41
translated nordic crime novelsTopic—5
callum dunmore poetryAuthor121

— means not in the top 50. Titles and authors are fictional.

Afterword

Get a plan like this

If readers find your work through someone else’s shopfront, Content Engine makes your own site the source they — and the search engines — trust first.

Finis

← Return to the library of case studies