How to find customers for a CNC machine shop

Who signs, and who actually decides

The design or mechanical engineer picks who gets the RFQ. They are choosing on whether you can hold the tolerance, whether you have done something like it, and increasingly on whether you will tell them their drawing has a problem rather than quoting it silently and making it wrong.

The buyer or procurement lead picks who wins from the list the engineer produced. They are choosing on price, lead time, certification and whether you are already an approved supplier, and they cannot add you to a list you were never on.

This is why capacity-list emails to procurement fail so consistently. Procurement does not go looking for new suppliers, they run a process against a list. The engineer is the one who can put you on it, and the engineer does not care about your spindle count until they care about your opinion on their part.

In smaller firms both jobs sit with one person, usually the technical director, and that is the easiest sale in this trade because the decision does not have to travel.

The five channels, in the order they pay

Ranked for a subcontract shop competing on capability and service rather than on being the cheapest, which is a contest you should not want to win.

1. 01

   Companies developing something new

   Best entry point, needs patience to convert

   A company designing a new product or machine needs parts made that nobody has made before, has no incumbent supplier for those parts specifically, and is talking to engineers rather than running a procurement process.

   That is the one moment your way in does not require displacing anyone. Prototype and low-volume work is where relationships in this trade actually start, and almost every long production contract began as somebody's awkward one-off.

   The evidence is public and mostly ignored: job ads for design and mechanical engineers, new product announcements, trade press, funding for hardware companies, exhibition stands. A company hiring three design engineers is developing something, whatever their website says.

   The discipline is not to price the prototype like a production run. You are buying a position on a list, and shops that treat the first job as the profit are the ones that never see the second.

2. 02

   Everything inside your freight radius

   Unglamorous, and it is where the margin lives

   Machining is a logistics business wearing an engineering hat. Freight cost, lead time and the ability to get somebody to walk a part across town are real advantages that survive a price comparison.

   So the highest-yield list you will ever build is every company within your practical delivery radius that makes a physical product. Not a sector, a circle. It is dull, it is finite, and most shops have never actually written it down.

   A customer two hours away who can drive to you when a part is wrong is worth substantially more than the same order from four countries away, and both of you know it the first time something goes wrong at three in the afternoon.

   The limit: your radius may not contain enough of the work you are good at. If that is true, the answer is specialising into something worth shipping further, not emailing everyone in Europe.

3. 03

   Certifications matched to a sector

   Expensive to acquire, then it does the selling

   Certification is a filter before it is a credential. AS9100 for aerospace, ISO 13485 for medical, IATF 16949 for automotive, plus ISO 9001 as the baseline nobody remarks on.

   What makes this a channel rather than an admin task is that it inverts the competition. Inside a certified sector you are competing against the handful of shops that also hold it, rather than against every shop with a milling machine and a lower rate.

   The corollary is that outreach into a certified sector without the certification is wasted effort, however good the parts are. Check what the sector requires before building the list, not after.

   If you hold something unusual, lead with it. It is the rare case where a credential genuinely belongs in a first line, because it is the thing that qualifies you to be read at all.

4. 04

   Reshoring and supply chain trouble

   Genuinely live in Europe, and it fades

   Companies that moved machining to distant suppliers have spent several years learning what a twelve week lead time costs when a design changes, and a real number of them are actively pulling work back.

   The signals are unusually explicit, because companies talk about it: supply chain resilience in an annual report, a stated shift to local sourcing, public complaints about lead times, or simply hiring a supply chain manager.

   This is fourth rather than first because it is a window rather than a permanent condition, and because the work that comes back is often the work that was hardest to make profitably in the first place. Quote it carefully.

5. 05

   Tenders and framework agreements

   Large, slow, and mostly not for small shops

   Public sector and large industrial frameworks put real volume in one place, and once you are on one the work is predictable in a way nothing else in this trade is.

   The cost is a bid process that assumes somebody has time to run it, and terms written for a supplier considerably larger than you. A shop of fifteen people can win these, but usually only in a genuine specialism where the field is thin.

   It is last because the time it takes is time not spent on the first two, and for most shops the first two pay sooner and better.

The signals worth watching

All visible from outside, and mostly published by the company itself. Manufacturing is unusually noisy about what it is building.

Hiring design or mechanical engineers

They are developing something, whatever the website says. New parts, no incumbent supplier, and a conversation with an engineer rather than a procurement process.

A new product, model or machine announced

Parts that did not exist last quarter. The clearest entry point there is, and it comes with a launch date attached.

New machinery, a new hall or capacity expansion

They are taking on more than they can make in-house. Overflow subcontract work usually follows within months.

Stated reshoring or supply chain trouble

Work coming back from far-away suppliers, and companies say so out loud in reports and interviews.

A sector certification you happen to hold

Not an event, a qualification. Inside a certified sector you compete with a handful of shops rather than all of them.

Anything that makes a physical product inside your delivery radius

Also not an event. Freight and lead time are real advantages, and this list is finite, which means it can actually be finished.

A directory of manufacturers tells you who owns a factory. A signal tells you who is drawing a part this quarter that nobody has made yet, and only one of those needs a machine shop right now.

The same signals, written as searches

This is what you would actually type into Wisemation. Not filters, not an industry code: it reads the meaning, which matters more here than in most trades because manufacturing classification codes are notoriously useless. Each one below turns a signal above into something you can run. Click any of them to run it on ten leads, free.

• “machine builders and equipment manufacturers in Denmark that posted a design or mechanical engineer job in the last three months”Run it →

  They are developing something. New parts, no incumbent supplier, and a posting from this quarter rather than an advert that closed last year.

• “medical device manufacturers in Bavaria and Baden-Württemberg that launched a new physical device this year”Run it →

  A launch date, real tolerances, and a sector where ISO 13485 thins the field dramatically. Something exists that had to be made.

• “manufacturers in northern Italy that announced a new plant or a factory expansion in the last year”Run it →

  Capacity committed before it can be staffed. The overflow goes out to subcontractors within months, and the announcement carries the date.

• “machine builders and equipment manufacturers within 200km of Eindhoven”Run it →

  No trigger at all, just your freight radius, and companies that buy machining rather than sell it. Finite, dull, and the list most shops have never written down.

The classification problem is worth naming. A shop that makes surgical instruments and one that makes agricultural fittings can sit under the same industry code, and a filter-based database cannot tell them apart. Describing what a company actually makes, in words, is the only thing that does.

What the first email actually says

The default machine shop email is a capacity list. Five-axis, wire EDM, tolerances to a hundredth, ISO 9001, quick turnaround. It fails for a specific reason: it answers a question the engineer has not asked yet, and it is indistinguishable from the last four that arrived.

What works is showing you understand what they make. Reference the product, the material, the thing that is hard about it. An engineer will read a stranger who has clearly looked at the part, and will delete a stranger who has clearly looked at nothing.

Then offer something engineers actually want, which is not a quote and not a capability brochure. It is a second opinion. Quote one part against their drawing, or say plainly that a feature will be expensive to hold and why, along with the change that would halve it.

That last move is the whole trade in one sentence. Telling an engineer their drawing will cost them money, before you have won anything, is the fastest credibility available to a machine shop and almost nobody does it.

It gives away a piece of manufacturing advice before asking for anything, which is what an engineer remembers. Not one word about spindles, tolerances or certifications, and the ask is a drawing rather than a meeting.

What this costs, done properly

All of it is doable by hand, and shop owners do it by hand, usually on a Sunday. This is what each step takes, and what is left of it when it runs itself.

You approve every lead and every message before anything leaves. What stays with you is the part that wins machining work anyway: looking at a drawing, knowing what it will actually cost to hold, and saying so.

What reliably does not work

Four things nearly every machine shop tries, usually all at once.

• The capacity list email

  Machines, axes, tolerances and certifications, sent to a generic inbox. It answers a question nobody asked, it is identical to every competitor's, and it reaches procurement, who are not looking for suppliers. Every part of it is aimed slightly wrong.

• Competing on hourly rate

  There is always a shop with an older machine, a cheaper region and less overhead. Winning on rate means winning the jobs nobody else wanted at a price that assumes nothing goes wrong. Compete on lead time, on tolerance you can genuinely hold, and on telling them when the drawing is the problem.

• Pricing the prototype as though it were the job

  The one-off is a position on an approved supplier list, not a profit centre. Shops that load the first order to cover setup routinely lose the production run to somebody who understood what was actually being bought.

• Emailing procurement to get on the list

  Procurement runs a process against a list they were given. They do not build it. The engineer builds it, and the engineer is reachable, interested in a good technical opinion, and almost never contacted by machine shops.