Sell cookbooks to libraries by finding the missing shelf.

Libraries are one of the cleanest channels for serious cookbooks because the buyer is not only buying recipes. A selector is buying local access, cultural memory, programming material, community relevance, and a durable public copy.

The system shape

Our method turns cookbook sales into a small evidence system. The human still owns taste, relationship, and final outreach. The system owns the repeated work: normalize the title, search catalogs, score collection fit, preserve evidence, and generate a next action.

1. Canonical title record. Lock the ISBN, subtitle, author, publisher, publication date, format, subject headings, and distributor path. For Soomaaliya, the ISBN is 9781958417836 and the publisher record identifies it as a Hardie Grant North America hardback about Somali food, memory, and migration.
2. Territory graph. Build a list of library systems, not just branches. The useful unit is the acquisition system: one selector or policy can cover many patron shelves.
3. Holdings detector. Search by ISBN first, then title variants such as "Soomaaliya" and the common misspelling "Soomaliyaa." Store the catalog URL, date checked, result, and a screenshot or permalink when the catalog exposes one.
4. Should-carry model. Score fit from public signals: metro scale, world cuisine depth, immigrant and refugee services, East African or Somali community relevance, food programming, and whether nearby peer systems already carry the title.
5. Outreach packet. If the title is missing and the score is high, create a short selector note: title metadata, community rationale, comparable holdings, review blurbs or publisher proof, and the library's own purchase suggestion path.
6. Feedback loop. Recheck flagged systems after a suggestion, purchase, review cycle, or seasonal programming moment. The record should move from "gap" to "on order," "found," or "deprioritized with reason."

What makes the system useful

The win is not a bigger spreadsheet. The win is an audit trail. A selector-facing campaign should know the difference between a library that already has a copy, a library with no visible record, and a library where the evidence is stale.

Pacific Northwest audit bench

This embedded bench is the first working version of the workflow. Open each catalog search, set the holding status, and the system flags any high-fit library where Soomaaliya has no visible local record. The page stores your audit state in this browser and can export the current list as CSV.

How this becomes a sales system

The same bench can become a production workflow: scheduled catalog checks, a source evidence bucket, selector contact enrichment, distributor availability, CRM tasks, and a purchase-status feedback loop. The outbound message should be generated only after the audit produces a real gap.

• Start with one book, one ISBN, and one territory before expanding the list.
• Use public catalog links and dated evidence so every flag can be audited.
• Rank libraries by fit before sending any selector-facing note.
• Send outreach through library-preferred purchase suggestion or vendor paths.
• Recheck after 30 days and retire flags when the catalog shows a record.

The next automation layer

For a real campaign, Chopshopr would wrap this in a bounded agent skill: input a title record and territory, fetch catalog/search evidence, produce a gap list, draft selector notes, and require human approval before any message leaves the system.