Knowledge

I went through around 2200 posts by people from manufacturing companies in Poland and in Germany, Austria and Switzerland. A report with eight problems and six steps for your plant - available in exchange for your e-mail.

After a meeting the salesperson records a voice note, the person in charge receives signals about clients and salespeople, and the sales plan is built from the history of results. A sales problem shows within days.

Labels, quality certificate, delivery note, certificates and declarations of conformity assembled by the system into one PDF the moment an order is released for shipment. The set is produced from the data and stays with the order.

Orders, delivery confirmations and delivery notes read by AI and written straight into the system's records. A human approves with one click, and every document can be found in seconds.

The RFQ is read by AI, matched against order history and turned into a draft quote. The pricing decision stays with a human, and the company answers in a day instead of a week.

Contact groups with intervals, tasks planned automatically and rep effectiveness calculated from what actually got done. The manager sees traffic lights, the CEO sees the whole sales structure.

A regular buyer fades for months before the thank-you call. A mechanism that reads orders, invoices and payments every day sees it earlier.

Demand calculated from accepted orders and a threshold tied to lead times: running out of material stops being a surprise.

A concept for a chat that checks production data itself before it asks people. The start of a series from a live project.

The eurozone PMI rose to 51.9 points in July. Three questions to check whether your processes can carry more orders.

Six in ten small firms run without ERP. What that costs on the shop floor, and where to start.

The chatbot has to identify itself as AI, realistic images need labelling, and edited text stays outside the obligation.

A customer requires batch tracking? You don't need a big system - you start with a single line and one scan.

Material, labour, machine, scrap, set-ups. Let the system work out the real unit cost from actual production data.

Jobs, assignment and routes in one place - an end to manual rota-building and double trips.

One place for the whole network's data and a standard enforced by the system - instead of gathering spreadsheets by phone.

E-invoicing, staff shortages, margin pressure. Offload the repetitive steps and serve more clients with the same team.

A third way between one monolith and a pile of separate spreadsheets: linked tools and a single source of truth.

Contract prices, volume discounts, prices with an expiry date - let the system work out the right price for the salesperson.

Cash gets stuck between finished work and the invoice. Make the invoice a by-product of closing the job.

The technology is rarely to blame. Three causes recur - and all can be avoided.

Lock-in comes down to one thing - who holds your data and the code. What to ask a supplier.

The first working project is usually a low-to-mid five-figure sum - depending on scope.

Not every problem needs a bespoke system - and not every one a subscription.

The first reflex is to replace everything at once. It is the most expensive move there is.

The best software comes from your own pain.

A clear answer to what the regulator expects of you and what you can do first.

The company grew faster than my knowledge of what was really happening inside it.

The problem starts when you have so many spreadsheets that, more and more often, you cannot tell which one is the real one.