Why do some phone tickets tap and others show a barcode?
Because they are different objects. A barcode is a picture, which any issuer can produce and any scanner can read. A tap needs a credential provisioned into the device chip, which needs an agreement with the platform, hardware that supports it, and readers at the gate.
Two different objects
Apple's wallet distinguishes them in its own vocabulary: an ordinary pass carries content the phone displays, and a secure element pass is a pass whose credential the device stores in its certified payment chip. Only the second answers a reader.
A barcode pass is, at bottom, an image with rules about when to show it. A tap credential is a piece of hardware provisioning, closer to a card being issued than to a file being downloaded.
That difference, not the operator's taste, is what decides which of the two a traveller ends up holding.
Who can issue which
Anybody can issue a barcode. It needs no relationship with the platform, no security certification and no hardware on the phone beyond a screen.
Issuing something that taps needs the platform to allow provisioning into the device, an arrangement with the wallet, and readers that speak the right application. That is why small operators and one off events use barcodes and why large transit systems use taps.
It is also why a ticket bought in a browser two minutes before boarding is almost always a barcode: provisioning is a process, and a picture is immediate.
What each is like in practice
A barcode needs a screen that is awake, bright enough and clean, held still for a camera or a scanner. It fails in direct sun, with a cracked screen and at an awkward angle, and it works on a printed sheet when the phone dies.
A tap needs nothing visible at all. It works with a dark screen where the platform allows it, in the rain, and in a second, which the comparison page sets out in general.
The gate hardware differs too: a scanner is a camera and a validator is a radio, and an operator that has one does not automatically have the other.
Which one an operator picks
Throughput decides it. A barrier that has to pass a crowd wants a tap, because a camera and a held phone add seconds per passenger, which the throughput page quantifies for a different setting and which applies here.
Reach decides the other way. A barcode works on every phone, on paper, by email and in a screenshot, and it needs no agreement with anybody.
Most systems end up with both: taps at the barriers they own and barcodes for everything sold outside them. That is not indecision, it is the two objects doing what each is good at.
Frequently asked questions
Why can I not add my ticket to the wallet as a tap?
Because a tap credential has to be provisioned into the device, which needs an agreement between the issuer and the platform. A barcode needs none.
Is a tap ticket safer than a barcode?
A barcode can be photographed and forwarded, and a provisioned credential cannot. What matters more is what the system checks at the gate.
Does a barcode work with a dead screen?
No, and that is its practical weakness. A printed copy is the fallback, which is one thing a tap ticket cannot offer.
Why do the same operator tickets differ?
Usually because they were sold through different channels. Provisioning takes time and a picture is immediate.