The business case for automation: payback, utilization risk, and people
How to judge a machine or software that replaces manual work: what it saves, how volume changes the answer, and how to handle the effect on staff. A worked example from a warehouse in the Netherlands.
Firm processes and online tests change from year to year and differ by office. Use this to prepare, and confirm the exact current steps on the firm's own careers page.
Key takeaways
- Automation swaps a cost that moves with volume (people paid by the hour) for a cost that stays the same (a machine you have already paid for).
- Common mistakes: Counting hours saved instead of costs that actually go away: if nobody leaves and nothing else is done with the time, nothing is saved.
- How many roles change, and when, compared with how many people leave each year anyway.
- Retraining: new jobs running, cleaning and maintaining the machines, and handling the exceptions the machine cannot.
Key idea
Automation swaps a cost that moves with volume (people paid by the hour) for a cost that stays the same (a machine you have already paid for). So it pays only if there is enough volume to keep the machine busy, for long enough.
Automation means machines or software doing work that people did by hand: robots welding car bodies, conveyor systems that bring shelves to pickers, software that reads invoices. The business case has three parts. What does it save each year, after its own running costs? How long until the savings repay the price (the payback period, covered in "Judging an investment" in Case math)? And what happens to the people whose work changes? Because automation turns variable costs into fixed costs, it raises operating leverage: profit grows faster when volume rises, and falls faster when it drops.
Bar chart: Industrial robots per 10,000 manufacturing employees, 2024. Values in robots per 10,000 employees. Republic of Korea: 1,220; Singapore: 818; Germany: 449; United States: 307; World average: 132; Mexico: 62.
Source: International Federation of Robotics, World Robotics 2025, press release of 8 April 2026.
So-what
Factories in different countries make very different choices between people and machines, so the same robot can be a clear yes in one place and a no in another.
The International Federation of Robotics links Korea's lead to its electronics and car industries, the two biggest buyers of industrial robots. The economics explain much of the rest of the spread: the saving from a robot is the labour it replaces, so where wages are lower, the same machine takes longer to pay back.
Worked case
Should a warehouse near Rotterdam automate its picking?
The prompt
An online retailer's warehouse near Rotterdam, the Netherlands, employs 50 pickers, each costing EUR 50,000 a year including taxes and benefits. A system that brings shelves to the pickers costs EUR 3 million installed, lasts about 10 years, and costs EUR 200,000 a year to run (maintenance, software, energy). At the planned volume it does the work of 16 pickers. Staff turnover is 20 percent a year. Should the retailer buy it? (Fictional company, illustrative figures.)
The structure
- Payback = one-off cost / (labour saved minus running cost)
- Yearly net saving at the planned volume
- Payback against a 10-year life
- Key: Volume risk: the saving if volume is lower
- People: can turnover absorb the change?
Working it through
1. Labour saved
16 pickers at EUR 50,000.
Labour saved a year (EUR):16 × 50,000 = 800,0002. Net saving
Minus EUR 200,000 a year to run the system.
Net saving a year (EUR):16 × 50,000 - 200,000 = 600,0003. Payback
EUR 3 million divided by EUR 600,000 a year.
Payback (years):3,000,000 ÷ 600,000 = 54. At half the volume
If orders come in at half the plan, the system replaces only 8 pickers, but still costs EUR 200,000 a year to run.
Payback at half volume (years):3,000,000 ÷ (8 × 50,000 - 200,000) = 155. Volume needed to pay back within its life
To repay EUR 3 million in 10 years it must save EUR 300,000 a year net, so EUR 500,000 of labour, which is 10 pickers out of the planned 16.
Share of planned volume needed (fraction):((3,000,000 ÷ 10 + 200,000) ÷ 50,000) ÷ 16 = 0.6256. People who leave anyway
20 percent of 50 pickers leave each year.
Pickers leaving a year:50 × 0.2 = 107. Time to absorb 16 roles
16 roles, with 10 people leaving a year on their own.
Years to absorb the change through turnover:16 ÷ 10 = 1.6
The recommendation
Buy it only if the retailer is confident volume will stay well above two thirds of plan. At the planned volume it saves EUR 600,000 a year and pays back in 5 years, well inside its 10-year life. But it is a bet on volume: at half the plan, payback stretches to 15 years, longer than the system lasts; it needs about 62.5 percent of planned volume just to repay its cost (before allowing for the time value of money, which makes the bar higher). On people, turnover of 10 pickers a year can absorb the 16 roles in about 1.6 years if hiring is paused, and some pickers can be retrained to run and maintain the system. Next step: test the volume forecast and agree the staffing plan with employee representatives before signing.
Risks: Volume falls or shifts to items the system cannot handle; Breakdowns stop picking entirely unless some manual capacity is kept; Pausing hiring for 18 months strains the remaining team in peak season.
People impact: what a strong answer covers
- How many roles change, and when, compared with how many people leave each year anyway.
- Retraining: new jobs running, cleaning and maintaining the machines, and handling the exceptions the machine cannot.
- The rules: in the European Union, an employer considering collective redundancies (many job cuts at once) must consult workers' representatives "in good time" and notify the public authority in writing, under Directive 98/59/EC. Other countries have their own rules, so ask.
- Quality and safety while the change happens, and the morale of the people who stay.
A food factory in Pune, India, can buy a packing robot for INR 40 lakh. It replaces 2 operators on each of 3 shifts, each costing INR 3 lakh a year, and costs INR 4 lakh a year to maintain. What is the payback in years? (Round to two decimal places.)
Counting hours saved instead of costs that actually go away: if nobody leaves and nothing else is done with the time, nothing is saved. Forgetting the machine's own running costs. Assuming the planned volume will arrive. Automating a step that is not the bottleneck, which cuts cost but adds no output. Leaving people out of the case until the end, when the plan depends on them.
What does automation usually do to a company's cost structure?
A robot was bought for a planned volume. Volume then falls 40 percent. What happens to its payback?
What is usually the first way to reduce the effect of automation on staff?
Companies automate as part of bigger cost programmes and digital and AI projects. The strategic moves page covers cost restructuring and AI and digital adoption, with when each pays.
Open the strategic movesSources for this lesson (3)
- Recognized public explanations of case-interview concepts and terms
- International Federation of Robotics, press release "Robot Density Surges in Europe, Asia, and Americas", 8 April 2026 (World Robotics 2025 report, data for 2024) (checked 2026-10-01)
- European Commission, Employment, Social Affairs and Inclusion: Collective redundancies (Directive 98/59/EC) (checked 2026-10-01)
My notes on this lesson
0 of 5,000 characters. Saves automatically.
Try the 4 remaining checks and drills above to complete this lesson (0 of 4 done).
Spotted something wrong or out of date? Report a mistake. We check every report and correct the page.