“And how much is this going to cost me?” is always the second question. The first one is “can this be automated?” — and the answer is almost always yes. If you’re not sure about your own process, our automation checker takes you through a quick test to find out. The trouble is that the honest answer to the second one is “it depends”, and “it depends” doesn’t help you decide anything.
So we’re going to do something unusual in this industry: give real ranges, explain what they depend on, and hand you a calculation you can do today on the back of a napkin to know whether automating a specific process pays off — before you ask anyone for a quote.
The short answer: real ranges
There’s no price for “automation”. There are processes, and each one falls into a level of complexity. These are the ranges we see in real projects for SMEs in Spain, once you strip out the hype:
- Simple automation (€1,500 – €5,000). A single system, clear rules, few exceptions. Moving data from one place to another, generating a recurring report, automatic alerts, reconciling two spreadsheets. Days or a few weeks of work.
- Medium automation (€5,000 – €20,000). Two or three systems that have to talk to each other (your ERP, email, a website), business logic with special cases, and usually some extraction of unstructured data — invoices, delivery notes, emails. This is the bulk of what an SME asks for.
- AI or fully custom automation (€20,000 and up). When you have to train or integrate a model (demand forecasting, classification, OCR on messy documents), connect to legacy systems, or build something that doesn’t exist off the shelf. Here the range opens up because the project looks more like custom development than “connecting pieces”.
A word of advice: be wary of anyone who quotes a fixed price without having looked at your processes. And be even warier of whoever quotes the cheapest. In automation, cheap is usually a demo that nobody uses three months later.
Why it varies so much: the four factors
If two automations that “sound the same” cost three times more one than the other, it’s almost always down to a combination of these four factors:
- Number of systems to integrate. An automation inside a single tool is cheap. The moment your ERP, your document manager and an external supplier all have to talk, the cost is set by the integration, not the logic.
- How structured your data is. A clean Excel file is a dream. A crooked scanned PDF from a supplier who changes the format every month is where the budget goes: that’s where OCR and AI come in, and that’s 80% of the effort.
- The volume of exceptions. The “normal case” gets automated in a day. What costs money is “except when the customer is in the Canary Islands”, “except at month-end close”, “except if field X is missing”. Each exception is code, tests and maintenance.
- Who maintains it afterwards. An automation is living software. If nobody owns it, it breaks silently the day a supplier renames a column. Maintenance (in-house or contracted) is part of the real cost, not an extra.
The back-of-the-napkin calculation: does it pay off?
Before asking for a single quote, do this calculation. You only need three honest numbers:
Annual saving = hours/week the process eats up
× real cost per hour of whoever does it
× 52 weeks
Payback (months) = project cost ÷ (annual saving ÷ 12)
The rule we use: if the project pays for itself in under 12 months, it’s almost always worth it. Between 12 and 24 months, it depends on how much the process hurts you for other reasons (errors, team stress, angry customers). More than 24 months, you’re probably automating the wrong thing.
If you’d rather not reach for a napkin, run the numbers in our savings calculator and you’ll get the payback in a few seconds.
One important detail: the saving is almost never just the time. It’s also the errors you stop making and the work you stop putting off. That’s harder to squeeze into the formula, but it’s usually half the value.
A real example (with numbers)
A distributor we work with was entering supplier invoices by hand: someone copied data from PDFs into an Excel and then into the ERP. Three days of one person’s week, plus stock errors that cost them returns.
The napkin calculation:
- Before: ~24 hours/week × ~€18/hour × 52 = ~€22,000/year in time alone, before counting the errors.
- Project: OCR over the invoices + automatic reconciliation against the ERP + demand forecasting. It fell into the medium-high range.
- Result: three days of work freed up each week, OCR with 98% accuracy and zero stock errors. The project paid for itself in well under a year.
We don’t tell it because the case is spectacular, but because it’s typical: the money wasn’t in the technology, it was in the time the team stopped losing.
Grants and digitalisation aid: count on them
If you’re an SME or a self-employed professional in Spain, a good part of these projects fall within the Kit Digital scheme and other digitalisation grants, which can cover a significant share of the investment. It doesn’t change whether a process is worth automating — the napkin calculation still rules — but it can improve the payback considerably. It’s worth checking what you’re entitled to before you start.
Three mistakes that blow up the cost (and how to avoid them)
- Automating a broken process. If the manual process is badly designed, automating it just makes it get things wrong faster. Fix the process first; then automate it.
- Trying to automate everything at once. 80% of the value is usually in 20% of the process. Automate that chunk, measure it, and decide the next one with data. A big all-in-one project is where the most money gets buried.
- Having no owner. If nobody on your team understands and watches the automation, it’s not an asset: it’s a time bomb. Handover and documentation aren’t optional.
In short
“How much does it cost?” is the wrong question to start with. The right one is “how much does it cost me today to do it by hand, and in how many months do I recover it?”. If you have those numbers, you already know more than most people asking for automation quotes — and you’ll be able to tell a good proposal from wishful thinking.
If you’d like, we’ll help you run that calculation on a specific process of yours. An hour of discovery is usually enough to tell you whether there’s a real saving there — or whether it’s not time yet. Let’s talk.