A yard app is not an office app
Picture a driver at seven in the morning, in a yard, in winter. He is wearing gloves, he is in a hurry, the phone is the company one with a screen that has seen better days, and the signal comes and goes because between the warehouses the network is what it is. He has to record a collection and photograph a document. This is the person the app must be designed for. Not the clerk at the desk with the big monitor, the mouse and the fibre connection: he already has the management system.
Most enterprise "field" apps fail exactly here. They are designed at a desk, with the head of someone who sits in an office, and out in the yard they die within weeks. The sign of death is always the same: paper comes back. The notepad, the sheet folded in four, the photo sent to a colleague on WhatsApp. People are not sabotaging anything: they are defending themselves against a tool that slows them down.
Why they fail
The reasons repeat. The app asks for a login with a complex password, every time, with gloves on. It demands connectivity for anything at all, and at the first tunnel or the first shielded warehouse it shows a spinner and then an error. It has the same dense screens as the management system, recompiled for a small display. And when something goes wrong it dumps the problem on the user: "operation failed, please retry". The driver does not retry: he goes back to paper, and he is right.
The principles that work
Offline first. The app has to work without a network as the normal condition: it records everything locally, queues it, and syncs on its own when the signal returns. The person using it should not even need to know whether there is coverage right now: tap, the work is recorded, done.
Three taps for the main operation. The gesture performed fifty times a day has to fit in two or three taps, with large targets, readable in sunlight and with gloves on. Everything else, the settings, the rare cases, can sit behind: the main screen belongs to today's work.
The camera is the keyboard. In the field you do not type: you frame. The document gets photographed and the system, behind the scenes, reads it and attaches it to the right record. A code gets scanned, not copied out. Every character typed in a yard is a potential error and ten seconds lost.
The login should not exist, or nearly: fingerprint or face, and you are in. Security moves to where it does not get in the way: to the registered device, to access that can be revoked remotely, to data that does not stay on the phone.
The app handles its own errors. If a submission fails, it retries by itself; if a value does not add up, it flags it to whoever can act on it, not to the driver who can do nothing about it. The rule is that a technical problem must never become a task for the person doing the work.
And the principle that contains all the others: you design by watching the real work. Half a day in the yard, next to the people, is worth more than ten requirements-gathering meetings. That is where you discover the glove, the sun on the screen, the place where the phone actually gets put down.
The objections I always hear
"Our workers will never use an app." They will use anything that saves them time. If the app removes a trip back to the office to hand in paperwork, they adopt it within a week. If it adds one more obligation to their load, they are right to refuse it, and the problem lies in the design.
"What if they lose the phone?" The data does not live on the phone: it lives in the company systems, and access is revoked remotely in a minute. A lost phone is a fifty-euro nuisance; a lost pad of documents is a serious problem.
"Wouldn't a rugged industrial terminal be better?" Sometimes yes, in extreme environments. But for most jobs the smartphone people already know how to use, inside a tough case, beats the thousand-euro handheld nobody likes and that stops everyone when it breaks.
What to take away
The test is simple and you run it before writing a line of code: can the main operation be completed with gloves on, in the sun, with no network, in under ten seconds? If not, the field will reject it. And the measure of success, after launch, is how much paper stops circulating. That one is easy to count: just look inside the door pockets.
Is the field in your company still running on paper? Tell me about it.