Library and Archive Labelling: Labels That Last Decades
Libraries and archives have a requirement almost no other sector shares: labels applied today must still be readable in thirty years, on items that are handled constantly and stored for decades.
Retention changes the media decision entirely
Everywhere else, a label lasts as long as the item is in the process. Here the item outlives the people who catalogued it.
Two consequences follow directly.
Direct thermal is unusable. It fades within months under heat and light, and a spine label on a shelf near a window will be illegible within a year. Thermal transfer is the only sensible choice.
Facestock and adhesive must age well. Paper labels yellow and become brittle. Adhesives can dry out and release, or worse, migrate into the item they are attached to. Our media guide covers facestock and ribbon selection generally, and archival work sits at the demanding end.
For genuinely archival material, adhesive interaction with the item is a conservation question rather than a labelling one, and worth taking advice on separately.
Handling is the wear mechanism
Unlike warehouse labels, which are read and forgotten, library labels are touched constantly.
Spine labels are rubbed by hands and neighbouring items every time something is shelved or removed. Over years that abrades the printing.
Resin ribbon on synthetic facestock resists abrasion far better than wax on paper. Many libraries also apply a protective covering over spine labels, which extends life considerably and is cheaper than relabelling a collection.
The failure is gradual and easy to miss: labels do not stop working one day, they become progressively harder to scan until staff start keying numbers manually β which is the point the problem finally becomes visible.
Small labels, dense codes
Spine labels are narrow, which constrains what fits.
Where a code must sit on a spine, it is small, which means 300 dpi printing to render it accurately and adequate scanner optics to read it. A 203 dpi printer producing a dense code at that size produces something that scans marginally rather than reliably.
Quiet zones matter particularly here, because the temptation on a narrow label is to use every millimetre. A code crowded against the edge will not decode however well it is printed, and this is one of the most common causes of labels that stop scanning. Our label design guide covers it.
Always include the number in human-readable text as well. When a label degrades or a scanner is unavailable, that text is what keeps the item identifiable β and over a thirty-year horizon, that will happen.
Scanning at the desk and in the stacks
Two quite different requirements.
At the issue desk, items are presented one after another. A presentation scanner reading continuously is faster than a handheld, because the operator moves the item past it with the hand already holding it rather than picking up a device, aiming and triggering. Across a busy period that changes the pace of the desk. Our front-of-store guide covers the same arithmetic.
In the stacks, for shelf-reading and stocktaking, a portable device is needed β and this is where the operator must see information back, not only capture codes. Confirming an item is in the right sequence requires seeing what the system expects, which is a mobile computer rather than a scanner.
Also consider that items are read at awkward angles on shelves, frequently in poor lighting, sometimes on curved or damaged spines. A 2D imager tolerates that considerably better than a laser.
Where RFID fits, and where it does not
Libraries are one of the sectors where RFID genuinely earns its place, and it is worth being precise about why.
Shelf reading. Sweeping a shelf and reading every item without removing any is transformative for a task that otherwise takes days. This is the strongest case there is.
Self-service issue and return, where multiple items are processed at once without presenting each barcode.
What it does not do is remove the barcode. Items still carry printed labels because the number must be human-readable, because a tag carries nothing visible, and because barcodes are read by anyone β including partner institutions and inter-library systems.
Most libraries that adopt RFID run both. Our RFID guide covers the general comparison, including the stray-read tuning that any deployment requires.
Specifying for a library or archive
How long must the label remain readable β and be honest, because the answer is usually decades. What is the smallest code and on what surface. Will items be handled constantly, which drives abrasion resistance and protective covering. Is the collection shelf-read frequently enough to justify RFID. And is there any conservation constraint on adhesives touching the items.
That first answer alone rules out direct thermal and most paper stock, which is where most labelling regret in this sector originates.
Common questions
Can I use direct thermal for spine labels?
No. Direct thermal fades within months under heat and light, and a spine label near a window will be illegible within a year. Thermal transfer onto synthetic stock with a resin ribbon is the only sensible choice for decade-scale retention.
Why do spine labels stop scanning gradually?
Abrasion from constant handling. Labels do not fail one day β they become progressively harder to scan until staff key numbers manually, which is when the problem becomes visible. A protective covering extends life considerably.
Why do narrow labels fail to scan?
Usually quiet zones. On a narrow spine label the temptation is to use every millimetre, and a code crowded against the edge will not decode however well it is printed. Small dense codes also need 300 dpi rather than 203.
Does RFID replace barcodes in libraries?
No, it supplements them. RFID is transformative for shelf reading and self-service, but items still carry printed labels because the number must be human-readable, a tag carries nothing visible, and barcodes can be read by partner institutions.
Presentation scanner or handheld at the desk?
Presentation, where items are processed one after another. The operator moves the item past a scanner already reading rather than picking up a device, aiming and triggering β one action instead of three.
Tell us how long labels must remain readable and we will specify media that lasts rather than media that prints.
