How a metalworking firm with 35 staff saves €17,000 a year

You can read plenty of general talk about digitalisation. This article covers one firm we have supported for 18 months, with the figures from before and after. We call it “Hartmann Metallbau”.

Note: The case in this article is a composite example from our work. Names and figures have been changed.

The business

• Sector: metal fabrication, mainly balconies and staircases
• Location: Chemnitz, Saxony
• Staff: 35, of whom 4 work in the office and 31 in production and installation
• Annual turnover before the project: around €4.2 million
• Founded: 1998, run by the second generation since 2019
• Our contact: Kerstin R., co-owner and head of finance and administration

Where they started (January 2024)

Kerstin approached me at an event run by the IHK Chemnitz (the regional chamber of commerce and industry). She summed up her problem in one sentence: “We’re growing, but every other step of growth costs me a bit of my sanity.” In detail, it looked like this.

Problem 1: stock kept in Excel

For years the stock records had lived in an Excel file with 17 tabs. Only the storekeeper maintained it. When he was on holiday, nobody could say how many steel sections were left. As a result, production stopped twice a quarter because material was missing on site. Each stoppage cost €400–800.

Problem 2: orders by email

Orders went to the three main suppliers by email, with a rough Excel export attached. Mistakes happened regularly: wrong quantities, wrong dimensions, wrong delivery dates. With one supplier alone, the firm averaged two complaints a month.

Problem 3: time sheets on paper

The fitters wrote their hours on slips of paper, and Kerstin typed them up in the evening. That meant 3–4 hours of copying every week. Job costing lagged behind accordingly. Kerstin only found out whether a job had made money 6–8 weeks later.

Problem 4: no warning of price changes

When a supplier raised prices, nobody noticed until they checked the invoice. In the worst case, three deliveries had already arrived before anyone saw the new price list.

What we did

Phase 1: process audit (week 1)

I spent three days on site at Hartmann. On day one I stood next to the storekeeper, on day two I sat with Kerstin in the office, and on day three I went out to a site with one of the fitters.

The result was three clearly separate building blocks that we could implement one after the other. Replacing all the software in one go was never on the table.

Phase 2: stock management (weeks 2–7)

We moved the Excel spreadsheet into a proper inventory management system. We chose WeClapp: not a big name, but GDPR-compliant, with servers in Germany and fair pricing. Along the way we:

• migrated only the 280 items the business actually uses, rather than all 800,
• set a minimum stock level for every item, working it out with the storekeeper rather than at a desk,
• set up an alert: when an item drops below its minimum level, the storekeeper gets an email and Kerstin sees it on her dashboard.

Phase 3: connecting the suppliers (weeks 6–9)

In parallel, we connected the three main suppliers via EDI (electronic data interchange) or structured email:

• The order from WeClapp goes into an email template. One click and it’s sent.
• The system compares the supplier’s order confirmation with the order and highlights any differences in quantity, price and date in colour.
• Once a month the system reads in the price lists. If a price rises by more than 3%, Kerstin gets an email.

Phase 4: time tracking by phone (weeks 8–12)

Every fitter got a simple app on their work phone. We built it with AppSheet; if you know the tool, you can put something like this together in 2 hours.

1. In the morning: tap “Clock in” and pick the project from the list.
2. At lunchtime: “Break”.
3. At the end of the day: “Clock out”, with a short note if needed.
4. The data goes straight into job costing.

The figures after 18 months

Metric	Before	After	Change
Production stoppages due to missing material	8 a year	0	−100%
Complaints to suppliers	24 a year	6	−75%
Time spent on job costing	3 hrs/week × 52 = 156 hrs	0.5 hrs/week × 52 = 26 hrs	−83%
Time until a job’s contribution margin is known	6–8 weeks	2 days	−96%
Average capital tied up in stock	€185,000	€115,000	−38%
Warning of price increases	after the fact	in advance	not measurable in figures
Kerstin’s working week	58 hrs	42 hrs	−28%

What that means in euros

Managing directors ask this first. Here is the calculation as we drew it up with Kerstin.

Savings per year

• Production stoppages avoided: €4,800 (8 × €600)
• Less effort on complaints: €3,600 (18 × €200 internal cost)
• Kerstin’s time freed up: 130 hrs × €45 = €5,850
• Less capital in stock (cost of capital and storage space): €2,800

Total: around €17,000 a year

Investment

• One-off implementation: €14,500. Half came back through the “Digital Jetzt” grant (a federal programme that part-funded digitalisation projects at small and mid-sized firms), so the business paid €7,250.
• Monthly support: €620 × 12 = €7,440
• WeClapp and AppSheet licences: around €1,800 a year

Running costs: around €9,240 a year

First-year balance (with grant)

• Costs: €7,250 + €9,240 = €16,490
• Savings: around €17,000

From month twelve onwards, the project pays for itself.

Balance from year two

• Costs: only the running €9,240 a year
• Savings: €17,000–20,000 a year, tending to rise as the team learns

That leaves around €8,000–11,000 a year on the plus side.

What Kerstin says today

“We had three other quotes before. One was twice the price, one was vague, and one was too cheap and sounded like a school project. Nulogic sat in the middle, but their consultant was the only one who had a coffee with us before making slides. That decided it. Today I ring them with technical questions as if they were my own IT department. And it costs a fraction of an in-house IT team.”

Kerstin R., co-owner of Hartmann Metallbau (name changed)

What you can take from this

1. One thing at a time. Three separate building blocks were manageable. One giant project would have failed. Ask every provider what they would explicitly not sell you.
2. Measure first. Without the before figures, we couldn’t prove today that it worked. Before you start, write down: how often does X happen? How long does Y take? How many errors are there?
3. Count the grant, but don’t rely on it. Digital Jetzt made the numbers easier. The project would have paid for itself without the subsidy too, after 18 months. If your project only adds up with a grant, take another close look at it.
4. Bring people on board early. Without the storekeeper, we would have set the wrong minimum stock levels. Without the fitter, the app would have been unusable in daily work. A software project always changes how a business works as well.

Is this relevant to you?

Stock in Excel, muddle with suppliers, paper time sheets: if any of that sounds familiar, you can very likely make similar savings. If you want to know exactly, book an on-site process audit: 1–3 days, from €1,500, fully credited against any follow-on project.

In an initial consultation, we’ll look together at where your business is losing time and money.