A royalty report tells you that 41 copies sold last month. It does not say that 19 of them followed a talk in Leeds, that the bookmark in the indie shop is being scanned twice a day, or that nobody in your home town has touched the book since launch. Scan data does.
What a scan records
When a reader scans a ScanSpine code, the redirect notes the time, the city and country (from the network, not GPS), the type of device, the site that referred them if any, and which code was scanned. If they go on to click a retailer link on your hosted page, that click is recorded too.
What it does not record: who the reader is. No names, no accounts, no cookies on their phone. IP addresses are hashed with a key that changes daily, which is enough to count unique scanners within a day and nothing more.
Questions scans can answer
Which placement works? Give each placement its own code: cover, bookmark, flyer, event banner. The dashboard shows scans per placement. The cover code almost always wins on volume; the surprises are in the others.
Did the event matter? A spike in a city on the day of a signing, then a tail of scans in the following week as people get home and open the book.
Where are readers you did not expect? A cluster of scans in a city you have never visited usually means a bookshop, a book club or a reviewer. Worth a look.
Do readers go on to buy? Clicks on retailer links, as a share of scans, show whether the hosted page converts. If scans are high and clicks are low, the page needs work, not the book.
Which retailer do readers prefer? The click breakdown often disagrees with where you thought your sales came from.
Reading the chart
Scans per day are noisy. Look at weeks, not days, and at changes against a baseline rather than absolute numbers. A book that gets three scans a day is a book in active circulation; a book that gets none for a month is sitting on shelves or in boxes.
Acting on it
- Double down on placements that scan; drop the ones that never do.
- Visit, or at least write to, the bookshop in the surprising city.
- Point the code at whatever the data says readers want: if clicks go to one retailer, consider sending them straight there during a promotion.
- Give the next print run a fresh placement code so you can tell editions apart.
The honest limits
Scans measure engagement with the physical book after it is in someone's hands. They do not count sales that never involved a scan, and a book bought as a gift and never opened leaves no trace. Treat scan data as a window onto reader behaviour that nothing else offers, not as a sales ledger.