Books and publishing
The link in chapter nine, ten years on.
Print copies can't be updated. Downloaded ebooks can't either. The companion link has to keep being right long after your site has been rebuilt twice.
The worked example
One title, one code, a decade of destinations.
Publishers' own advice is "use a dynamic code and check every link before each printing." That's a manual job we do every five minutes.
What changes for you
Three things you stop doing.
Checking links before every printing
Every destination is probed every five minutes, for the life of the book. If one dies, you get one email and the code shows a hold page instead of a 404 until it's back.
Freezing your website for the backlist
Rebuild the site whenever you like. Each title's code is one destination to update, not a thousand copies to worry about.
Guessing which chapters get scanned
Scans are counted by code and by day, people only; link previews and crawlers are kept apart. A chapter code tells you which chapter readers reach for.
A thousand titles
One code per title. Your own domain on every one.
Print under a hostname you own, like go.yourpress.com/9780000000000. Readable slugs are a recovery feature: that address can be rebuilt on any web server in an afternoon, and if we ever go away you repoint one DNS record and every book keeps working. The whole catalogue also downloads as a file any web server can serve.
Bring a spreadsheet of ISBNs and destinations and import it; or reserve a block of codes for the season's list and assign destinations as each title's page exists. The record of what each code pointed to, and when, is kept for as long as the code is.