Blog

Before you budget for another machine

October 23, 2026

Before you budget for another machine article graphic

It is budget season. In shops across the country, next year’s list is taking shape, and somewhere near the top sits a machine. A new press brake. A five-axis mill. A bigger laser.

The case for the machine is easy to write. You can quote cycle times and capacity gains. You can walk a customer past it. The payback math fits on one page, and everyone in the room understands it.

This piece is about the line item that never makes the list.

The capacity you already pay for

Walk the same shop on a normal morning and watch where the hours go. Someone is building a quote from scratch for a part the shop has made before. Someone is chasing the status of a late job across three conversations. Someone is re-keying an order into a second system, and someone else is assembling Friday’s report by hand.

None of those hours appear on a production report. All of them are paid for. That is the odd thing about budget season. A shop will finance a six-figure machine to gain capacity while an office full of capacity goes to searching, chasing and re-typing. (Our research on the hidden hours of shop-floor admin puts numbers on how much.)

There is a reason the machine always wins the argument. Decades of investment have gone into the production side of the building, in better machines, better tooling and better automation. The admin around the work has had nothing like that attention, so its cost has never been priced. It hides inside wages.

A side-by-side worth running

Picture Meridian Welding, a 45-person fabricator, building its 2027 budget. (Meridian is a fictional shop; the example is illustrative.)

Option one is the machine: a known price, known financing, and a payback case built on more parts per hour, which pays off if the work keeps coming.

Option two is less glamorous: connect quoting, orders, scheduling and reporting in one system, so the hours now spent searching and re-keying go back into quoting more work and shipping on time.

Option one adds capacity to make. Option two recovers capacity Meridian already pays for every month, usually for far less than the machine costs. Neither option is wrong. The problem is that most shops only ever cost the first one.

What it looks like when it is real

Volo Aero MRO, an FAA repair station running Tangle, put AI to work on one job: building repair orders from the customer’s PDF. Andrew Walmsley, their President, shared the result on LinkedIn.

"15 minutes to 3 minutes: That's how much faster we will now process a typical three-line repair order at Volo Aero MRO, thanks to AI integrated into our Tangle.io ERP system."

No new equipment. One task. Read the Volo Aero MRO story.

That is the shape of the office opportunity. Pick one repetitive, error-prone piece of admin and take it off someone’s plate. If you want a plan for choosing that first task, our free playbook, The AI-Powered Job Shop, walks through it.

Run your own numbers

The honest way to settle “machine or office” is arithmetic, not instinct. We built the free Five Levers ROI Model for exactly this. You enter your own figures and it estimates a year of value across the five places an ERP earns its keep: quoting, scrap, on-time delivery, inventory, and admin hours.

Run the Five Levers ROI Model

If the model says the machine wins, buy the machine. Some years it will. But the cheapest capacity in your building is usually the hours you are already paying for, and this is the season to price them.


By Tangle Software Inc. Tangle is the world’s first self-customizing ERP for manufacturers, with Milo wired into every customer’s instance. tangle.io.