A €300,000/yr business that runs on 4 human hours a week.

He started manual on purpose. His view was that if you automate a process you don't understand yet, you automate the wrong thing.

So for five years the whole business ran on seven Google Forms, one Google Sheet, Stripe and Gmail.

• Data in. Customers typed into one of seven forms, one per process, covering company onboarding, driver and vehicle onboarding, plate changes, offboarding, renewals, objections and an eligibility check. A form accepts whatever you type into it.
• Checking. He read the sheet line by line and corrected errors by hand.
• Invoicing. At the end of each month he counted each company's vehicles and adjusted the quantity in Stripe himself.
• Customer changes. He processed every driver and vehicle change as it came in.
• Printing. He emailed a CSV to a printing company.
• Final check. When the cards came back from the printer, he checked and laminated every one at his kitchen table.

Click to expand

The old system. Seven Google Forms and one master sheet, running a six-figure business.Close

Mistakes turned into physical objects in the post. A typo in a licence plate meant a card that didn't match the vehicle, which is the one thing the card exists to do. A wrong address meant a card posted somewhere that didn't exist. He once overcharged a customer because his vehicle count was wrong, which he said felt stupid.

He added what validation Google Forms allows, like forcing phone numbers to be 10 digits. It helped a little. It could not tell him whether a plate belonged to a real vehicle or an address to a real building.

Checking also got less reliable as the volume grew.

"I spent a lot of time checking all my data line by line, correcting it by hand. But after a certain volume it becomes harder, and you're getting so used to checking everything's right that you're going too quickly yourself."

Being the validation layer also meant being available. He answered every request the day it came in, seven days a week. On one family holiday he told his family he'd stay at the house for the morning, then worked the afternoon instead.

He knew the setup was wrong and had a backlog of things to fix. He tried to rebuild in a no-code tool, did two courses, and struggled for months. Learning a new platform at ten at night after a full day of work didn't work for him.

Then Claude Code came out at the start of 2026. He installed it on a Thursday, and by the end of that day it had built the things he'd been stuck on for months. Within a week, everything on his backlog board was working code.

• Data in. A custom app replaces the forms. Licence plates are checked against the RDW, the Dutch roads authority, so when you type a plate the app shows the make and model for the customer to confirm. A typo returns the wrong car or no car at all, and the customer corrects it themselves. Addresses are checked against a Dutch address database, and formats are enforced on entry.
• Checking. This is gone as a separate job, because the system refuses data it can't verify.
• Invoicing. Invoices are generated in the app from the live vehicle count, and payments come in through Mollie and are matched automatically.
• Customer changes. Each company now gets a dashboard to manage its own drivers and vehicles, and drivers get their own dashboard too if their company allows it, with permissions set per role. His largest client arrived with 800 vehicles and 800 drivers, so he built a CSV import for them. Local authorities have their own view of which plates hold an exemption.
• Printing. He chose a new print supplier, ADDI, because it has an API. Data goes out, and status comes back as printed, in progress or dispatched.
• Final check. This one is gone too, because the new supplier laminates the cards as well.

Click to expand

A client's dashboard, with demo data. They add vehicles, approve driver requests and see their whole fleet without emailing him.Close

Claude Code makes it much faster to add new features. A client asked him to stop drivers editing their own vehicle details, because the company runs a leased fleet. The feature didn't exist, and he said he could turn it on for them anyway. He put the phone down, prompted Claude Code, and called back fifteen minutes later to say it was done.

Build time. He started on a Thursday, and everything on his backlog was working code within a week. Migrating five years of spreadsheet data and testing took another six weeks before a single customer touched it.

Difficulty, hard. This is a full custom application with logins, payments, per-role permissions, two government API connections and a supplier integration, and it holds other people's personal data. It is not a weekend project.

What he already had matters more than the difficulty rating. He'd spent five years running the process by hand, so he understood every step of it, and he had an existing paying customer base, so the rebuild had a known spec instead of a guess. His country has free public APIs for plates and addresses, which not every country does, and he switched print suppliers to get one willing to offer an API.

• Operations take about 4 hours a week, down from about 8. These are his own estimates from the call, nobody was timing anything in either period.
• Customer questions take almost none of that time, he says.
• He took four weeks off this summer. He did about 3 calls a week and roughly an hour of email, he didn't open Claude Code once, and anything that occurred to him went on a list for when he got back.
• Live since the start of 2026 and still running, on a customer base that includes an 800-vehicle fleet.
• Run cost about €354 a month, plus a Claude Max plan. The figure comes from his own billing. He thought it sounded high and offered to check it, so treat it as an upper estimate.

The rest of his working time now goes on improving the platform and growing the business.

The revenue picture. His revenue has grown significantly since the switch, to around €300,000 a year. The business was growing anyway, so the platform doesn't get all the credit. It does get some though, because a professional self-service product helped him win bigger, more established clients, and the operations time it freed up now goes on sales and marketing instead of admin. Sales is now the only part of the business that still needs him, and when he paused outbound over his summer holiday, growth paused with him.

For five years he checked everything by hand because mistakes came through constantly, and laminating every card at his kitchen table was his final look before anything reached a customer.

The rebuild removed the reason for the checks rather than just the checks. A wrong plate or address can't get into the system anymore, because everything is validated at entry against official databases and the customer confirms their own details on screen. The check moved to the front of the process, where a computer does it better than he ever could by eye.

He didn't take that on faith either. He kept having the printed cards sent to him at first, and only stopped once the new system had run long enough that he stopped finding mistakes. It has held up since going live in January.