One app for the shop floor: how I help a company move its internal processes off paper
It started with lunch. A manufacturing company had tried an app for meal orders that never quite worked, went back to the phone for the rest, and after enough problems ended up on paper. So we're building a new one. What we're really building, though, is the place where everything else will go next: announcements people have to sign, alerts when a machine stops, and the statistics nobody has time to collect by hand.
The company I’m writing about didn’t start from paper. It started from an app. A combined system was supposed to handle meal orders, and it never quite did everything people needed, so the missing parts went back to the phone: call the kitchen, add a portion, change an order. After enough of that, people gave up on the app altogether and went back to paper lists.
I find that story more useful than any success story, because it’s how most internal software fails. Not with a crash. People just quietly stop using it and go back to whatever worked before.
So the first job wasn’t technical. It was making sure the new app does all of the everyday thing, with no “call us for the rest”, so nobody has a reason to reach for the phone or the pen again.
I’m not going to pretend lunch is a big deal on its own. But it has one property that makes it the perfect first step: everyone uses it, every day. Once every employee has the app on their phone, knows how to sign in and trusts that it works, everything after it gets easier. That trust is the hard part, and at this company it had already been lost once.
What we’re building first
Two systems, working together. Canteen runs the kitchen side: menus, orders, cut-off times, company subsidies, and a counter terminal that keeps working when the network doesn’t. FirmFlow is the employee app and the layer around it: who works here, which department they belong to, what their role allows them to see, and a record of who did what.
On the phone, employees see the week’s menu, order lunch in one tap and show a barcode at the counter instead of a plastic card. It signs in with Face ID and it still shows today’s order in a basement canteen with no signal. The app is being prepared for its first release to the company’s people now.
The part that matters for everything else is invisible. FirmFlow doesn’t do lunch. It knows who you are and hands you a short-lived pass to the canteen system, and the canteen system does the rest. The same pattern works for any other service a company already uses.
What comes next
This is the roadmap we’re working towards, not features that exist today.
Announcements people actually have to confirm. A safety rule changes, a shift pattern moves, a new policy needs signing. Today that’s a printed sheet and a signature list on a clipboard, and a week later nobody knows who hasn’t signed. In the app, the announcement goes to exactly the people it concerns, each of them confirms it on their phone, and the manager sees who is still missing.
Machine faults, reported where they happen. An operator taps what stopped and why, maybe adds a photo, and the right technician gets a notification straight away instead of hearing about it at the end of the shift. Once those reports exist as data rather than notes, the failure statistics come for free: which machine stops most often, at what time, for how long.
That’s where we’re heading first, and the list isn’t closed. Once people trust the app with lunch, every paper form and every “just call me” in the company is a candidate for the next module, and the company decides which one hurts most.
If you want to do this in your company
I’d start the same way every time, and I’d resist the urge to do it all at once.
Pick the first process by how often people touch it, not by how important it sounds. Daily and company-wide beats strategic and monthly. Lunch, shift swaps, leave requests, safety sign-offs are all good candidates.
Connect to what you already have. Most companies don’t need a new canteen system, a new HR tool or a new maintenance database. They need one app in front of them, one login, and permissions that make sense. Replacing a working system is expensive. Connecting to it usually isn’t.
Plan for the shop floor, not the office. Shared phones or personal ones, gloves, halls with no signal, people who don’t have a company email. These decide whether the app gets used, much more than how it looks.
Measure something from day one. How long orders took before, how many sign-offs were missing, how long a machine waited for a technician. Without a number from before, you’ll never be able to show what changed.
And keep the first version small. One process that everybody uses is worth more than ten that half-work.
A quick test for your own processes
A process probably belongs in the app if most of these are true: it happens at least weekly, it involves people who don’t sit at a computer, it currently lives on paper, in a spreadsheet or in somebody’s head, someone spends time chasing other people about it, and nobody can tell you how it went last month.
If that sounds like something in your company, I’d be glad to look at it with you. I’ll tell you which process I’d start with, and honestly whether an app is worth it at all. Here is how I build these apps.