Four weeks ago, a client sent me an invoice. Not for our services — but his internal calculation of what the manual work before automation had been costing him. EUR 3,840 per month. For a single process. One employee who spent 3-4 hours daily copying data back and forth between three systems.
His automation costs with us: EUR 200 per month. Break-even: month 1. Not month 6, not month 12 — month 1.
And that's exactly where the problem lies with most automation discussions: people talk about technology, about tools, about artificial intelligence — but not about the only number that matters. The return on investment. This article doesn't deliver theory. It delivers a concrete calculator, three example calculations, and an honest analysis of when automation pays off — and when it's burning money.
My thesis, which I repeat in every initial consultation: Any automation that doesn't pay for itself within three months is the wrong automation. Not because faster payback isn't possible. But because there's almost always a process that pays off faster — and you should tackle that one first.
The formula is simple. Yet most companies calculate it wrong — because they only include the obvious costs.
The basic formula:
Manual costs per year = Hours per week x 52 weeks x Hourly rate (incl. employer costs)
Automation costs per year = One-time costs + (Monthly license x 12) + Maintenance + Training
ROI = (Manual costs - Automation costs) / Automation costs x 100
Break-even = Total automation costs / Monthly savings
The hourly rate isn't gross salary divided by working hours. It's the full employer costs: gross plus social contributions plus proportional office rent plus IT costs plus vacation days plus sick days. For an employee with EUR 3,500 gross, the real hourly rate in Germany is EUR 42-55 — depending on location and industry. Many companies calculate with EUR 25-30 and then wonder why automation supposedly doesn't pay off.
The following calculation is your guide. Insert your own numbers and calculate along.
Inputs for your ROI calculation:
Calculate manual costs: Annual manual costs = Hours x Hourly rate x Weeks. Example: 15 x 48 x 48 = EUR 34,560
Determine automation costs:
Automation costs Year 1 = Setup + Training + (License x 12) + Maintenance. Example: 2,000 + 300 + 2,400 + 500 = EUR 5,200
Automation costs from Year 2 = (License x 12) + Maintenance. Example: 2,400 + 500 = EUR 2,900
ROI result:
Annual savings (Year 1) = Manual costs - Auto costs Y1. Example: 34,560 - 5,200 = EUR 29,360
ROI Year 1 = Savings / Auto costs x 100. Example: 29,360 / 5,200 x 100 = 564%
Monthly savings = (Manual costs / 12) - Monthly license. Example: 2,880 - 200 = EUR 2,680
Break-even (months) = (Setup + Training) / Monthly savings. Example: 2,300 / 2,680 = 0.9 months
Read that last line again. Break-even after 0.9 months. Under one month. This isn't an outlier — it's the norm when the right process is automated.
Three concrete scenarios, based on real client projects. The numbers are realistic, not optimistic.
Scenario 1: Small business (5-15 employees)
Automated process: Invoice receipt + account assignment. Previous time required: 8 hrs/week. Hourly rate (employer costs): EUR 42. Annual manual costs: 8 x 42 x 48 = EUR 16,128. Setup (one-time): EUR 1,500. Monthly costs: EUR 150. Annual maintenance: EUR 300. Automation costs Year 1: EUR 3,600. Savings Year 1: EUR 12,528. ROI Year 1: 348%. Break-even: 1.3 months.
Scenario 2: Mid-sized company (20-80 employees)
Automated process: Order processing + warehouse management. Previous time required: 18 hrs/week (split across 2 employees). Hourly rate (employer costs): EUR 48. Annual manual costs: 18 x 48 x 48 = EUR 41,472. Setup (one-time): EUR 4,000. Monthly costs: EUR 350. Annual maintenance: EUR 800. Training: EUR 500. Automation costs Year 1: EUR 9,500. Savings Year 1: EUR 31,972. ROI Year 1: 336%. Break-even: 1.5 months.
Scenario 3: Large company (100+ employees)
Automated process: Multi-process: Invoices + Orders + Reporting. Previous time required: 45 hrs/week (split across 5 employees). Hourly rate (employer costs): EUR 55. Annual manual costs: 45 x 55 x 48 = EUR 118,800. Setup (one-time): EUR 12,000. Monthly costs: EUR 900. Annual maintenance: EUR 2,000. Training: EUR 1,500. Automation costs Year 1: EUR 26,300. Savings Year 1: EUR 92,500. ROI Year 1: 352%. Break-even: 1.6 months.
The pattern is clear: regardless of company size, break-even for well-chosen processes falls between 1 and 2 months. First-year ROI consistently exceeds 300%. From Year 2, it climbs further because one-time setup costs drop away.
To put this in context: our pricing model at Exasync is based on the principle that automation should cost about 25% of an employee's salary. Not 80%, not 50% — a quarter. In this range, the financial risk for the client is minimal, but the leverage is enormous. A concrete example: for our client Welzhofer, automation saves 3-4 hours daily. That translates to roughly EUR 800 in monthly savings. The cost: about EUR 200 per month. ROI was positive after the first month.
This is where it gets honest. Because most automation providers only show you the pretty calculation: setup costs versus saved hours. Reality has more line items. If you don't account for these, your ROI looks better on paper than in practice.
1. Maintenance and updates (5-15% of setup costs annually)
Every automation needs maintenance. APIs change. ERP updates shift database fields. A supplier changes their order format. Budget 5-15% of initial setup costs as annual maintenance. On a EUR 4,000 setup, that's EUR 200-600 per year. Not a fortune, but it belongs in the calculation.
2. Training costs (one-time + ongoing)
Your employees need to learn how to work with the automated process. Not because the automation is complicated, but because the workflow changes. Instead of typing data, they're now reviewing results and handling exceptions. That's a different activity and requires a different mindset. Budget 2-4 hours of training per affected employee, plus a refresher after 6 months.
3. Downtime costs during transition
During the transition phase — typically 1-2 weeks — old and new processes run in parallel. This costs time because employees are running both simultaneously. Plan for 20-30% productivity loss during this phase. For a team of 3 people at 40 hours per week, that's 24-36 lost hours. One-time, but real.
4. Opportunity costs from wrong automation choices
The most expensive hidden item. If you automate the wrong process, you don't just lose the investment — you also lose the time during which you could have automated the right process. For a failed project with a 6-month timeline and EUR 15,000 in costs, the opportunity costs are often higher than the direct losses.
5. Integration costs in legacy IT landscapes
In mid-market companies, clean IT architecture is rare. ERP from 2015, CRM from 2020, Excel and email in between. When automation needs to connect these systems, interface costs arise that are often missing from initial estimates. Rule of thumb: for each additional system that needs to be connected, budget an extra EUR 500-1,500.
Complete hidden cost calculation example:
Setup: EUR 4,000 (obvious). Monthly license Year 1: EUR 3,600 (obvious). Maintenance/Updates: EUR 500 (hidden). Training (4 employees x 3 hrs x EUR 48): EUR 576 (hidden). Parallel operation (2 weeks, 30% loss): EUR 1,440 (hidden). Additional integration (1 system): EUR 800 (hidden).
Obvious total: EUR 7,600. Hidden total: EUR 3,316. Real total costs Year 1: EUR 10,916.
In this example, hidden costs make up 30% of total costs. This doesn't fundamentally change the ROI — it stays positive, and significantly so. But it shifts break-even from 1.5 to about 2.2 months. Still under 3 months. Still a good investment. But realistic.
This is the part most providers leave out. We don't. Because there are clear cases where automation is the wrong approach.
Case 1: The process runs less than weekly. A process that takes 2 hours once a month costs you EUR 1,152 per year at EUR 48 per hour. Save 70% of that through automation? That's EUR 806 in savings. If automation has setup costs of EUR 3,000 plus EUR 100 monthly, it never pays off. Do the math before you start.
Case 2: The process changes constantly. Automation needs stability. If a process changes every 6-8 weeks — new rules, new exceptions, new systems — then adjustment costs eat up the ROI. Our rule of thumb: a process should run unchanged for at least 6 months before it's automated.
Case 3: The hourly rate is low and volume is small. At an hourly rate of EUR 25 and 3 hours per week, annual manual costs are EUR 3,600. An automation with EUR 2,000 setup and EUR 150 monthly costs EUR 3,800 in the first year — more than the manual work. It simply doesn't pay off here.
Case 4: The human value IS the process. Customer consulting, negotiations, creative work. When the human isn't the bottleneck but the product, automation is counterproductive. At most, automate the preparation and follow-up.
Case 5: Error tolerance is zero. If an automation error has severe legal or financial consequences and there's no human review step, think three times. Automation with human-in-the-loop? Yes. Full automation for critical processes without a review step? No.
The honest summary: if the scoring value (frequency x time investment x hourly rate) yields less than EUR 5,000 in annual manual costs, it gets tight. Not impossible, but tight. The best ROIs come from processes with manual costs above EUR 15,000 per year. Break-even there is almost always under 3 months.
The following analysis shows month by month when a typical automation pays for itself. Starting values: EUR 4,000 setup, EUR 200 monthly, EUR 2,880 monthly manual costs.
Month 0 (Setup): Cumulative investment: EUR 4,000. Cumulative savings: EUR 0. Net: -EUR 4,000.
Month 1: Cumulative investment: EUR 4,200. Cumulative savings: EUR 2,880. Net: -EUR 1,320.
Month 2: Cumulative investment: EUR 4,400. Cumulative savings: EUR 5,760. Net: +EUR 1,360.
Month 3: Cumulative investment: EUR 4,600. Cumulative savings: EUR 8,640. Net: +EUR 4,040.
Month 6: Cumulative investment: EUR 5,200. Cumulative savings: EUR 17,280. Net: +EUR 12,080.
Month 12: Cumulative investment: EUR 6,400. Cumulative savings: EUR 34,560. Net: +EUR 28,160.
Month 24: Cumulative investment: EUR 8,800. Cumulative savings: EUR 69,120. Net: +EUR 60,320.
Break-even falls between month 1 and 2. From month 2, the automation generates EUR 2,680 net every month. After 2 years, net profit totals over EUR 60,000 — on an initial investment of EUR 4,000. This isn't marketing wishful thinking. It's math.
But: these numbers only hold when the right process is automated. If you automate a process with EUR 500 in monthly manual costs, the table looks completely different — and break-even is at 12-18 months. That's why process selection is more important than tool selection.
Over the past months, we've seen enough projects to identify a clear pattern. The difference between fast and slow payback isn't about the tool, the budget, or the industry. It comes down to three factors:
Factor 1: Process frequency. Daily processes pay back 5x faster than monthly ones. This sounds obvious but gets constantly ignored. Companies love to automate the elaborate quarterly report instead of daily data entry. The quarterly report has more glamour. Data entry has more ROI.
Factor 2: Clarity of requirements. If the process is clearly documented before automation — including all exceptions and special rules — the project runs in 2-3 weeks. If the service provider first has to figure out how the process works, it takes 6-8 weeks. The documentation phase before automation saves more money than any tool.
Factor 3: Number of connected systems. A process within a single system (e.g., email automation in Outlook) can be automated in days. A process connecting three systems (ERP + CRM + email) needs weeks for integration. Each additional system extends time-to-value. That's why you should start with single-system automations.
The combination of these three factors explains why some projects pay off in 4 weeks and others in 12 months. It's not random. It's selection.
Here's a concrete 3-step plan you can implement this week. No external tool, no consultant, no budget needed.
Step 1: Time log for 5 working days. Ask 3-5 employees from different departments to note how they spend their time for a week. Not in detail — rough blocks are enough. Emails: 2 hours. Data entry: 1.5 hours. Report creation: 1 hour. At the end of the week, you have a list of processes with actual time requirements.
Step 2: ROI calculation for the top 3 time sinks. Take the three processes with the highest weekly time investment. Calculate the annual costs (hours x hourly rate x 48 weeks). Compare with realistic automation costs. If break-even is under 3 months, you have a candidate.
Step 3: Launch a pilot. Not three processes at once. One. The one with the best ROI. Start small, see results fast, then scale. The first success is the most important one because it convinces the team that automation works and doesn't threaten jobs.
If you've done the math and the ROI checks out, but internal implementation isn't feasible: talk to us. At Exasync, we implement automations that pay for themselves within 1-3 months. Not because we're particularly cheap, but because we only take on projects where the math works out. The pricing model is simple: automation costs about 25% of an employee's salary. If that doesn't clearly outweigh the manual work, we'll honestly tell you it's not worth it.
More about how AI is concretely used in businesses can be found in our article AI for businesses. If you want to know which business processes are best suited for automation, read our guide to business process optimization. And if you're looking for a partner who covers the entire process from analysis to implementation: How an AI automation agency works.
The numbers don't lie. The question isn't whether automation is worth it for your business. The question is which process has the highest ROI — and whether you start this quarter or spend next year watching your competitor save money.
