In a small manufacturing or service company, a "time and attendance system" still often means paper sign-in sheets and an HR person who spends three days a month copying them into Excel. It works — until it doesn't. And it usually stops working abruptly: in a dispute over overtime, during a labour inspection, or when the company grows from 15 to 40 people and the monthly payroll close starts spilling into a full week.
There's a lot of noise around time and attendance. Hardware vendors promise "cloud T&A for €7 per person", mobile app vendors talk about "100% paperless", biometric terminal makers scare you with fraud stories. The truth is much simpler and less sexy: T&A in an SME pays off in a few very concrete scenarios, and outside them it's a needless cost. This piece is about telling one from the other.
42 people, 4-day payroll close, 6 disputes a month
At one of our clients — a 42-person metalworking company (two production shifts + office + warehouse) — the HR lead described the starting point as: "monthly payroll close takes me four days, three of which are copying". Paper attendance sheets arrived from four shift supervisors, with different delays and different quality. Overtime was calculated by hand. Business trips and customer visits sat in email threads. A dispute of "I stayed longer and wasn't paid" happened on average six times a month, and each one ended in a quiet settlement because you couldn't formally prove anything.
The company didn't want an "enterprise system". They wanted the payroll close to take half a day, not four. And they wanted to stop arguing about overtime.
The rollout: two RFID terminals at the shop-floor entrance and at the office entrance (separate zones — because production runs on fixed hours, the office runs on flexi-time), integration with Comarch Optima Kadry i Płace (Polish payroll suite) via the terminal vendor's API, and a simple mobile app for five sales reps on the road. One-off cost: around 24 000 PLN (two terminals, cards, deployment, integration, training). Running cost: around 350 PLN a month (SaaS licences plus support).
Three months later: the payroll close shrank from four days to half a day — meaning 14–16 hours of the HR lead's time back every month. Overtime disputes dropped to zero (there's nothing to argue about; the data is there). Shift supervisors stopped writing sheets — they look at the Optima report. Payback: around 10 months, counting only HR time at fully-loaded cost. Previously unbilled overtime — now properly accounted for — added the same again, but we don't count that as savings, because formally that was always money owed to the workers.
This isn't a story about "AI revolutionising HR". It's a story of common sense: where people physically walk onto a shop floor and off it, a T&A terminal is cheaper than an HR person with a pen.
When T&A actually pays off
Not in every company. Three scenarios where we see a real return:
- Production or warehouse with at least 15 people physically on the floor. A terminal plus integration with your ERP's timesheet module starts paying back when you have at least 15–20 shift workers and even one supervisor who currently loses an hour a day to this.
- Shift work with overtime. If night shifts, weekends and holidays happen regularly, T&A isn't a luxury — it's legal protection. One labour-court case over unpaid overtime costs more than five years of the system's subscription.
- Field work: installation, service, transport. Terminals are useless here, but a mobile app with geolocation replacing a paper trip sheet can pay back in three months — it closes the gap between "I was at the customer for four hours" and "you were there for two".
When T&A is money burned
And three scenarios where it's better to hold off:
- A pure office of 8–15 people, flexi hours. A terminal at the door will start fights ("the card didn't read, I was there until 5") and won't add anything. In a small office, the payroll close takes 2 hours, not 4 days. There's nothing to optimise.
- A company where trust is the cultural foundation. If the team has worked together for years and overtime is done "for a colleague", T&A can break that social contract in a week. Sometimes the cost of the resulting churn beats the savings.
- A company that doesn't yet have contracts and schedules in order. If you don't know who is on which working-time regime, who is task-based and who is standard, and what the overtime rules are — T&A won't fix that. It will automate the mess.
The order is always the same: sort out the rules and the schedules first, then deploy the system. Not the other way around.
Three technologies, three real drawbacks
The Polish T&A market is unnecessarily complicated. You need to know three variants and their trade-offs — the rest is marketing.
- RFID card or fob. The standard on shop floors. Cheap (~2 500–4 000 PLN per terminal), reliable, low-conflict. Downside: employees lend each other cards. "Clock me in, I'll be there in ten". The scale of this grows with turnover. The fix is a camera at the terminal and a rulebook. Not biometrics.
- Biometrics (fingerprint or face). Solves the card problem but introduces new ones. Legally — under GDPR — biometric data is a special category and requires an explicit legal basis and a DPIA. In practice a good lawyer will talk you out of this in any company under 100 people unless you have a genuinely real reason (e.g. access to high-value areas). Don't buy biometrics out of inertia.
- Mobile app with geolocation. For sales reps, field service, installers. Great for logging customer visits and routes. Downside: workers dislike it, and they're legally right to — the employer has to formally regulate GPS monitoring in the work rules (per the Polish Labour Code and GDPR). Deploying without that = risk of a labour inspection.
In most manufacturing SMEs the right pick is RFID at the door plus a mobile app for field workers. Biometrics only when you truly need it.
ERP integration: where projects break
T&A on its own is simple. The complicated part is integration with the payroll system — Comarch Optima Kadry i Płace, enova365, Sage Symfonia, Płatnik. That's where most mistakes happen, so a few specifics.
- Don't buy a terminal whose vendor doesn't have a ready connector to your payroll system. "We'll give you a CSV file for manual import" in practice means your HR person will still be keying things in every month. Confirm the Optima / enova / Symfonia connector exists before you buy, not after.
- Agree who is the source of truth for time worked. If the terminal says "8:07" and HR corrects it to "8:00", you'll never earn the team's trust. Agree the rounding rule (e.g. 5-minute rounding — up on entry, down on exit) in writing and put it into the terminal's configuration, not into the HR person's head.
- Plan the exception handling. Absences, holidays, business trips, remote work, broken cards. Each of these cases must have a defined home in the system — otherwise it drifts back to paper, this time dressed up as an email to HR.
- Pilot with one department for a month before rolling out to the whole company. One shift, two weeks running in parallel (paper and terminal), a month to reconcile the differences. Without that, the first full payroll month will be a disaster.
We've seen T&A projects where the hardware was installed in a week and the ERP integration dragged on for six months because nobody agreed on the rounding rules. The risk lives in the process layer, not the hardware layer.
Labour law and GDPR — what you can't skip
We aren't a law firm, but three things show up in every project and it helps to know them before you pay a deposit:
- The work rules ("regulamin pracy") must describe the T&A system. How time is tracked, what happens if someone doesn't clock in, and how employees access their own data. Without updating this document, the labour inspectorate can strike down a "late arrival" penalty based on a terminal report.
- GPS monitoring in the mobile app is a separate consent. The Labour Code requires informing employees and adding monitoring to the work rules, at least two weeks before it goes live. You can't just install the app and roll.
- Data retention. T&A records are kept for the working-time record period (currently 10 years in Poland, changed a few times). A terminal that keeps them "forever" doesn't comply with GDPR's minimisation principle. Check this in the vendor contract.
Nobody tells you this on a sales slide. We do, because we've seen three projects stall in month two because someone forgot about the work rules.
What to do this week
If reading this felt like a description of your company, five things to do before you even talk to a vendor:
- Count how many hours a month your time tracking costs. HR, shift supervisors, the owner. All in. If it's less than 20 hours a month, T&A probably won't pay back.
- Count how many overtime disputes there were in the last 12 months. Each one is at least two hours of several people's time, plus damage to the working atmosphere. That's real money.
- Check your current payroll system. Comarch Optima, enova365, Symfonia, R2Płatnik — each has its own integration standard and its own T&A add-on price. This decides which terminal vendor you pick, not the other way around.
- Ask the shift supervisors what annoys them most about the current process. 80% of the time the answer is "having to collect it all on paper myself". You just found your project ally.
- Update the work rules before, not after deployment. That's two hours with a lawyer. Without it, the whole project is legally suspended in mid-air.
T&A in an SME is not a project that changes a company. It's a project that closes one specific, expensive problem — and it closes it cheaply, if done with a head on your shoulders. Don't sell it to everyone; don't praise it where it doesn't pay; don't buy biometrics to brag on LinkedIn.
If you want to check whether it will pay off at your place, half an hour of conversation is usually enough to work it out.
T&A won't optimise a company that doesn't have order in its schedules yet. It will automate the mess. Start with the rules, then the hardware.