US Manufacturing
How US Procurement Managers Find New Manufacturing Suppliers

TL;DR
Procurement managers rarely start with a cold Google search for 'CNC machine shop.' They search for the part, the process, the tolerance, or the certification. They also ask ChatGPT, check ThomasNet, scan LinkedIn for who a competitor uses, and ask engineering for a referral. Shops that win RFQs show up across all four paths, not just one.
Quick answers
- Where do procurement managers actually look for new suppliers?
- Google and AI search for the specific part or process, ThomasNet, LinkedIn (searching who a peer company already uses), industry directories, and referrals from engineering or existing vendors.
- Do procurement teams still use ThomasNet?
- Yes, mostly for discovery and shortlisting, not as the only source. A free, fully filled-out listing still surfaces in category searches.
- Is ChatGPT part of supplier discovery now?
- Increasingly yes, for early-stage shortlisting. Buyers ask it to list shops that do a specific process and region, then verify on the shop's own site.
- What search terms do buyers actually type?
- Process plus tolerance plus material plus region, such as '5-axis CNC titanium aerospace Ohio', not generic terms like 'machine shop near me.'
Most manufacturing marketing advice still treats supplier discovery like it is 2015: buyer searches "CNC machine shop," sees ten paid ads, calls three. That is not how a procurement manager or sourcing engineer at a $5M to $50M OEM actually works in 2026. The search is narrower, the trust bar is higher, and there are more starting points than there used to be.
How does a procurement manager actually start looking for a supplier?
Atomic answer: they start from the part, not from the category. A sourcing engineer with a drawing in hand searches for the process, tolerance, material, and often region, such as "Swiss turning stainless 0.0005 inch tolerance medical," not "precision machining company." Shops that only optimize a homepage for "CNC machining company" miss this traffic entirely.
The four real discovery paths
Shops we audit assume there is one path (usually "our website should rank"). In practice, RFQs arrive through a mix of four paths, and most shops are only visible on one or two.
| Path | What the buyer is doing | What gets a shop found |
|---|---|---|
| Search (Google, Bing) | Typing process + material + tolerance + region | Capability pages per part family, not a generic services page |
| AI search (ChatGPT, Perplexity) | Asking for a shortlist of shops that do X in region Y | Clear, citable capability statements in plain text, not buried in PDFs |
| Directories (ThomasNet, industry associations) | Filtering by category, certification, location | Fully completed free listing, real photos, certifications entered |
| Referral / network (LinkedIn, peer recommendation) | Asking "who do you use for X" or checking who a competitor lists as a vendor | Active LinkedIn presence, case studies, being named by existing customers |
None of these paths is optional if the goal is a steady RFQ pipeline. A shop that only shows up in directories gets commodity price-shopping RFQs. A shop that only shows up in search but has no LinkedIn presence loses the "who else has vetted this shop" trust check that most procurement managers now run before sending a drawing.
What buyers type into search (and why generic terms fail)
Procurement managers searching for a new supplier almost never search the way a shop's homepage is written. Common real patterns:
- "5-axis CNC aluminum aerospace bracket Ohio"
- "contract manufacturer AS9100 prototype to production"
- "sheet metal fabrication tube laser Michigan"
- "injection molding medical grade resin ISO 13485"
Compare that to a typical shop homepage headline: "Your partner in precision manufacturing solutions." None of the words a buyer actually searches appear on the page. This is the single most common gap we see when auditing US manufacturer websites: rich shop-floor capability, but a website written in marketing language instead of buyer language. Our guide on capability pages that convert covers how to fix this page by page.
AI search is now a real discovery channel
Buyers, especially younger sourcing engineers, increasingly type a full question into ChatGPT or Perplexity before opening Google: "which US shops do 5-axis titanium machining for aerospace." The tools respond with named shops pulled from indexed, plainly stated capability content. If a shop's tolerances, materials, and certifications are only described in a PDF brochure or a slider image, none of that text is readable by these tools. See AI search visibility for US manufacturers for the mechanics of getting cited, and how AI is changing supplier discovery for the buyer-side view.
Directories still matter, but only if complete
ThomasNet, IndustryNet, and sector-specific directories remain a discovery source, particularly for buyers doing a first-pass shortlist search by category and certification filter. The free listing tier still gets crawled and still surfaces in category searches. Shops that fill every field, certifications, equipment list, real photos of parts, materials handled, tend to outperform shops that paid for promotion but left the profile half-filled. Our full breakdown is in ThomasNet alternatives for US manufacturers.
Referral discovery still runs through a website check
Even when a buyer is referred by a colleague or a machine tool rep, the next step is almost always the same: open the shop's website to confirm the referral before picking up the phone. If the site does not clearly state tolerances, materials, certifications, and location, that referral goes cold, not because the shop lacked capability, but because the site failed to confirm it fast enough. What happens next in that visit, and where shops lose the buyer even after being found, is covered in why your competitor gets the RFQ you should be getting.
A quick self-check
Run through this before assuming discovery is the problem:
- Does your homepage or a capability page contain the exact process + material + tolerance phrases a buyer would search?
- Is your ThomasNet (or equivalent) listing 100% complete with real photos?
- Does your LinkedIn company page show recent activity and name-able work?
- If someone pastes your capabilities into ChatGPT, would it find plain text to cite, or only an image or PDF?
If any answer is no, that is a visibility gap costing RFQs before a buyer ever reaches your RFQ form. Our free Buyer Reach Audit checks all four in about 60 seconds.
CTA
If you are not sure which of these four paths is leaking RFQs, run the free Buyer Reach Audit on your own site. For a deeper look at your full funnel from search to signed RFQ, book a 30-minute manufacturing growth audit.
60-second live scan
See where your buyers are searching - right now.
Run your buyer-intent keywords through live Google. Get a competitor map and ranking gaps in 60 seconds. Free, no pitch.
Run my free Buyer Reach AuditPrefer a working session?
Book a 30-min call with our team.
No pitch. Walk away with a written action plan - whether you hire us or not.
Book a strategy callFor US manufacturers
Tired of ThomasNet junk leads?
Run a free 60-second audit of where your buyers are searching in the US. Get a real diagnosis — no pitch.
Questions about this topic
Should a small shop bother with SEO if most business comes from referrals?
Yes. Referrals get a buyer to your name, but they still search your site to verify capability before sending an RFQ. If the site does not confirm what they were told, the referral dies quietly.
How much does a Google Business Profile matter for a B2B shop?
Less than for local retail, but it still shows up for 'near me' style searches and gives a fast trust signal: reviews, photos, verified address.
Does a shop need to be on every directory?
No. Being complete on two or three (ThomasNet, your own capability pages, LinkedIn) beats being thin on ten.
How do buyers verify a supplier once they've found a name?
They check the website for tolerances, certifications, materials, and equipment list, then look for a case study or gallery of similar parts before requesting a quote.
Related field guides