We won the order. The warranty can figure itself out.

Everybody in the market buys electronics, steel, cables, components and labour. Somehow our quotation still manages to arrive suspiciously near the bottom.

Maybe we are simply more efficient.

Or maybe some competitors have this strange habit of charging enough money to support the equipment after they sell it.

We see margin differently. Margin creates distance between us and the lowest bid. Warranty reserves make the quotation ugly. Spare-parts stock is money sitting on a shelf doing absolutely nothing.

So we trim.

A warranty sounds cheap until somebody uses it.

Then suddenly we need an engineer. A replacement board. International shipping. Diagnostics. Maybe flights and hotels. Maybe several days of technical work for a customer whose invoice was paid two years ago.

Proper manufacturers expect this and keep money available for it. We expect the equipment not to fail.

This produces excellent financial results right up until the equipment develops an opinion.

And if a claim arrives at an inconvenient moment, our technical diagnosis becomes extremely motivated to discover installation error, misuse, environmental conditions or anything else that comes with somebody else’s invoice.

Customers occasionally expect us to keep important spare parts available. Control boards. Modules. Contactors. Displays. Components that can put an expensive machine back into operation quickly.

Unfortunately, stocking parts requires money.

A spare board sitting unused on our shelf is terrible for cash flow. The same board sitting on somebody else’s shelf is called supplier inventory.

So when something breaks, we place an order too.

This creates a wonderfully equal experience: the customer waits for us, and we wait for our supplier.

Customers sometimes misunderstand what 24/7 support means.

It means you can send the email twenty-four hours a day, seven days a week.

Resolution is a separate product category.

We can acknowledge the ticket, ask for photographs, request logs, schedule another call and copy increasingly senior people into the email. None of this requires us to maintain a large dedicated support organisation.

If the problem actually requires an engineer on-site, then the economics become more interesting. Flights cost money. Hotels cost money. Engineers apparently expect salaries while travelling.

So every field-service visit becomes a small internal referendum on how badly we want to honour the warranty.

The beautiful thing about underfunding future obligations is that the future eventually arrives with new customers.

A warranty claim from an old project requires money today. A new order also brings money today. We enjoy these coincidences.

Suddenly we can buy the missing part, send something urgently or free up engineering time to fix equipment whose original margin disappeared long ago.

Of course the new machine now has its own future warranty obligation.

That is tomorrow’s problem.