It is Thursday afternoon. The production supervisor has spent three hours on next month's rota in Excel, because two people are on sick leave and one comes back from holiday mid-week. On Monday it turns out that an operator coming off a night shift is booked for the morning shift. The fix goes out on a WhatsApp group, and nobody updates the spreadsheet.
This post will let you decide which of three options fits your company: Excel with built-in rules, an off-the-shelf rota app, or a tool of your own. Three numbers drive the choice: how many people work shifts, how many rules the supervisor keeps in their head, and how many times a month the rota changes.
What an Excel rota really costs
In 20–50 person companies with shift work we typically see 4–8 hours of a supervisor's time to build the monthly rota, plus another 2–4 hours a week on corrections. That cost is easy to count. The harder one to count is the errors a spreadsheet does not catch.
Polish labour law sets specific requirements here. Under the Polish Labour Code (Kodeks pracy), an employee is entitled to at least 11 hours of uninterrupted rest in every 24-hour period (Art. 132 § 1) and at least 35 hours in every week (Art. 133 § 1). The employer must hand over the work schedule at least one week before the period it covers begins (Art. 129 § 3). Source (in Polish): Art. 129 of the Labour Code, Art. 132, Art. 133.
A plain spreadsheet checks none of this. The supervisor checks it, by eye, on Thursday at 5 pm. The most common mistake we see is employees swapping shifts on a handshake, after which someone has an 8-hour break instead of 11. The second problem: after a week the spreadsheet and reality drift apart, so payroll calculates overtime from a rota that no longer applies. If you also run time and attendance, we covered this in our post on time and attendance tracking in SMEs: a system will automate the mess if the rota is not in order first.
Three routes and who each one is for
1. Excel with rules. The same spreadsheet, but with conditional formatting that highlights rest shorter than 11 hours, an understaffed shift and exceeded working hours. Plus one rule: every change goes into the spreadsheet, not into WhatsApp. You can do it yourself in 1–3 days or outsource it for PLN 2–5k (about EUR 500–1,200). It is enough for a dozen or so people, a fixed rotation and one supervisor who builds the rota.
2. An off-the-shelf rota app. The Polish market has several SaaS apps for shift planning, billed as a monthly subscription per employee. They give you what Excel cannot: employees see the rota on their phone, submit availability and swap requests, and the app enforces rest periods. Rollout usually takes 1–2 weeks, mostly spent onboarding people and entering rules. Price the subscription over 3 years for your headcount and compare it with option 3.
3. Your own tool. It makes sense when the rota depends on things an off-the-shelf app does not know: who is qualified on which machine, who holds a valid forklift medical, how many people a specific production order from the ERP needs. Typically PLN 25–60k (about EUR 6,000–14,000) and 4–10 weeks for a web application (in our case usually .NET) integrated with Comarch Optima or enova365 (popular Polish ERP systems) or with time-clock terminals. Then maintenance: budget 10–15% of the build cost per year.
A simple rule: if the rules fit on one sheet of paper and do not involve qualifications or orders, stay with option 1 or 2.
Two examples: one that worked, one that overpaid
One of our clients, a 38-person manufacturer of cardboard packaging, runs three shifts on 9 machines. The rota depended on who could operate which machine, and only the supervisor knew that. The off-the-shelf apps the client tested could not handle a qualification matrix tied to the production plan in Comarch Optima. We built a simple web app: qualification matrix, rest-period rules, phone view and an export of hours to payroll. The project took 6 weeks and cost PLN 38k (about EUR 9,000). The supervisor's time on the rota dropped from about 5 to 1 hour a week, and violations of the 11-hour rest rule, previously 3–4 a month, stopped occurring. What we would do differently: write the qualification matrix down before the project starts, because agreeing it with the foremen took two of the six weeks.
In another case we were called in after the rollout. A 24-person distribution centre had ordered a custom tool from its previous vendor for PLN 45k (about EUR 10,500). After three months the warehouse manager went back to Excel. The reason was not technical: most rota changes were agreed by phone the day before, and the tool required every swap to be approved in an admin panel. The problem was the swap process, not missing software. An off-the-shelf app where employees request swaps on their phone would have cost a fraction of that. Lesson: before you order anything, count how many rota changes happen after it is published and who approves them today.
When your own tool makes no sense
Custom rota software is the most expensive of the three options and the easiest one to overinvest in. Do not build it if:
- fewer than 15 people work shifts and the rotation is fixed (e.g. a week of mornings, a week of afternoons). Excel with rules will last for years.
- the rota depends only on people's availability, not on qualifications, machines or orders. That is exactly what off-the-shelf apps do.
- nobody in the company will own the tool after rollout. Rules change (a new machine, a new shift pattern, a change in work regulations), and every change means several hours of developer time.
- the problem is discipline, not tooling. If swaps happen on a handshake, no software will fix it without a decision that only what is in the rota counts.
To be fair, here is what you give up with an off-the-shelf app: your rota data sits with the vendor, the payroll export often comes in one rigid format, and changing rules depends on what the vendor anticipated. At 20 people that is usually an acceptable trade-off. At 45 people across three halls it may not be. We write more about when a spreadsheet stops being enough in when Excel becomes the system.
What to do this week
- Count how many hours the supervisor spent building and correcting the rota last month. Ask them directly; do not estimate for them.
- Take last month's rota and compare it with the attendance list or time-clock data. Note how many changes happened after publication and how many never made it into the spreadsheet.
- Manually check 10 random "night shift, next shift" pairs for 11 hours of rest.
- Write down on one sheet of paper every rule the supervisor applies when building the rota: minimum staffing, qualifications, who cannot work with whom, preferences. If they fit on one sheet and do not involve machines or orders, option 3 is out.
- Add one conditional format to the current spreadsheet that highlights a break shorter than 11 hours. It is an hour of work and the first test of whether the problem exists at all.
You choose a rota tool by three numbers: how many people work shifts, how many rules the supervisor keeps in their head, and how many times a month the rota changes. Your own software only makes sense when the rota depends on qualifications, machines or orders.