What is an acceptable failure rate in a batch?
Whatever you agreed in writing, and the number matters less than knowing which number you are talking about. A supplier rate is measured at the factory before anything you do. Yours includes printing, handling and installation, so the two are never comparable.
Two different numbers
A supplier quotes a rate measured at the end of their line, on tags that have not been printed, laminated, transported by you, handled by your team or applied to anything.
Your observed rate includes every one of those. It is a rate for your whole process, not for the product.
Comparing the two directly produces an argument neither side can win. A batch arriving within specification and failing at five times that rate in your hands is not necessarily a supplier problem.
Separating them is the first step, and it requires reading tags on arrival before you do anything to them.
Measuring your own
Read every tag on arrival, or a defined sample if the quantity makes that impractical. Record the count. That is your incoming rate.
Read every tag again after printing, laminating or converting. The difference is your process rate, and it is usually larger than the incoming one.
Read every tag after writing, which is the step most people already do. The difference again is your writing rate.
Three numbers, three causes, three different fixes. A single overall figure tells you something is wrong and nothing about where.
Agreeing one with a supplier
Put a figure in the order along with how it is measured and when. A rate with no measurement point is not an agreement.
Specify that it applies on arrival, before any processing, since that is the only condition both sides can verify.
Agree what happens when it is exceeded: replacement, credit, or a returned batch. Deciding that in advance turns a dispute into a procedure.
And agree a sample size, because a rate measured on ten tags and a rate measured on a thousand are different claims with different confidence.
What the rate should be
Low enough that individual failures are an inconvenience rather than a pattern, and high enough to be honest. Any physical product has a failure rate, and a supplier claiming none is describing their marketing rather than their line.
What matters far more than the number is where the tags are going. A rate that is unremarkable for stickers on a desk is unacceptable for tags being installed on a roof, because the cost of each failure is the site visit rather than the tag.
So set the requirement from the cost of replacement, not from an industry figure. That is the calculation that makes a rate meaningful, and it is the one that justifies paying for individually tested tags where the installation is expensive.
Then verify every tag before installation rather than relying on any agreed rate. Catching a dead tag on a bench costs seconds and catching it on a wall costs a return trip.
Frequently asked questions
What rate is normal?
It varies by product, process and supplier enough that a borrowed figure is not useful. Measure your own incoming rate over a few batches and use that as your baseline.
Should I test every tag?
Before installation, yes, wherever reaching the tag again is expensive. On a desk with easy replacement, sampling is reasonable.
The supplier says zero failures. Is that credible?
As a claim about individually tested tags leaving the line it can be. As a claim about what arrives after transport and handling it is not.
Who is responsible if tags fail after printing?
You are, unless you agreed otherwise, since printing is your process. This is why measuring the incoming rate separately matters.