Back to Blog
  • AI Search Visibility
  • Entity SEO

Marketing for Manufacturers in Windsor-Essex: What ChatGPT Cites for Suppliers

Hayden Bond12 min read
Marketing for Manufacturers in Windsor-Essex: What ChatGPT Cites for Suppliers
ChatGPT cited a shop's own website in 79 of 90 answers when we asked it, from Canada, where to find machine shops, tool and die, mold makers, coaters and fabricators around Windsor-Essex. Most of those citations went to a home page. So marketing for manufacturers here starts with a plain question: what does your home page tell a buyer, and ChatGPT, about the work you do?

88%

Share of ChatGPT's answers about Windsor-Essex suppliers that cited at least one shop's own website. Shops' own sites made up 58% of everything it cited.

Source: Plate Lunch Collective field study, October 20, 2026
We asked ChatGPT 30 questions a plant buyer might ask when looking for a supplier near Windsor, Ontario and the Detroit border, three times each. Then we searched the shops it named most on Google, the way a buyer checks a shop before sending an RFQ. It's the kind of search our manufacturing work is built around.

Marketing for manufacturers starts with what your home page says

Shops' own websites were cited in 79 of the 90 answers and made up 408 of all 704 citations (58%). When ChatGPT cited a shop's site, it was usually the front door. Of those 408 citations:
  • 263 (64%) were home pages
  • 52 (13%) were capabilities, services or equipment pages
  • 4 (1%) were quality or certification pages
PDFs were 13 of all 704 citations, about 2%. We can only say what got cited. The study can't tell whether a shop's certification page was missing or just wasn't read.
We've been telling people that the AI answer is the new first impression, and the site is where the buyer goes for proof. In my view, a shop's website matters to buyers because that's where the proof lives. Before sending an RFQ, a buyer is vetting the shop. What certifications does the shop hold? What does it specialize in? How do procurement and replenishment work? How far is it from the major shipping hubs?
I'd say many shops lean on words like "the best," "cutting edge" or "industry leading." Words like that don't explain capabilities, certifications or specialized production lines and machinery. Manufacturing sites need plain language that scopes what the shop does, where people and AI can find it easily. None of it should be hidden in a downloadable document or a hard-to-reach file.
Put that next to the numbers. When ChatGPT cited a shop's site, it cited the home page 64% of the time and a certification page 1% of the time. So I'd put the proof where it gets read, starting with the home page.
The answers placed suppliers in Windsor, Ontario itself (75 answers), Oldcastle (48), Windsor-Essex (39) and Tecumseh (26). Oldcastle and Tecumseh came up often enough that a shop there should say where it is on its site.

Industry networks and directories get cited next to the shops

Shops weren't the only sources. An industry association or network was cited in 35 of 90 answers (39%). A business directory, such as Yellow Pages or Thomasnet, was cited in 29 (32%). A map listing was cited in 29 (32%). After Google Maps, the single most-cited site was an online directory of Ontario machine shops, in 18 answers (20%). That's more than any one shop's website.

What ChatGPT cited in answers about Windsor-Essex suppliers

A shop's own website

88%

Industry association or network

39%

Business directory

32%

Map listing

32%

Plate Lunch Collective field study, October 20, 2026. 30 questions, three ChatGPT runs each, from Canada, logged out.

In my view, this comes down to specificity and trust. Associations and directories have their own curated set of data, organized around what buyers check. They show business details in a standard layout, often with schema markup: the labels in a page's code that tell a machine which line is the address and which is the service. Simply adding schema to your own site won't help on its own. But it presents a cleaner record. And answers aren't built from one source. Models put them together from many, and trusted directories and associations are among them. We counted what got cited, not why.
We did check the schema point on pages ChatGPT cited often. Of the 10 network and directory pages we could read, 6 showed a structured list of businesses, and three of those were from one network. Of the 12 shop pages we could read, 4 had no structured data, 7 had only the basic markup a website plugin adds, and 1 described the business itself with its address. That's a small sample, and it doesn't show that schema is why a page got cited.
The advice questions went somewhere else. On the five questions about sourcing across the border and how to find or vet a supplier, 7 of 15 answers named a shop. Government or economic-development sources were cited in 8 of those 15 answers, against 7 of the other 75. When we asked how to find a reliable contract manufacturer in Ontario, all three answers pointed to associations, directories and government programs, and none named a shop. When we asked how to vet a new machining supplier, ChatGPT answered without searching the web, citing a source or naming a shop, in all three runs. So when a buyer asks how to find a supplier, the first stop ChatGPT suggests may be a list your shop is or isn't on.

When a buyer searches your shop's name, do they find you?

Next we took the 40 shops ChatGPT named most often and searched each one on Google Canada by name. That's the check a buyer might run after getting a name from an AI assistant. We've written about that split between AI making the suggestion and Google being where people check it.
For 8 of the 40 shops (20%), the results were mostly about a different company with a similar name. In one, Google's AI Overview described a different tool and die shop in another Ontario city. If the buyer doesn't notice, they're reading about someone else. We wrote more about how AI decides which business you are.
Across the 40 name searches:
  • The shop's own site came first in 18 and was somewhere in the top 10 in 21. For 3 shops, no official site turned up at all.
  • Google's business panel showed up for 33. Of those, 30 showed reviews and 31 linked a website.
  • A business directory was in the top 10 for 23 shops. LinkedIn, Facebook or another social site was there for 35.
  • An AI Overview appeared for 19, and cited the shop's own site in 9.
Just because your website has your name and logo on it doesn't mean buyers and search platforms know you are who you say you are. I'd say other sources need to vouch for you and verify your claims, so your record is clean and you can be named clearly. If a business with a similar-sounding name has a stronger record online, it's the one that comes up. That's my read on why the mix-ups happen; the study shows that they happened, not why. Checking that the sources AI reads agree on who your shop is and what it does is the work on our entity SEO page.

ChatGPT's list of shops changed from one run to the next

ChatGPT named 288 different companies across the 90 answers. We counted each company once for each question it was named for, which gave 396 question-and-company pairs. Only 21% came back in all three runs of the same question. Another 27% came back in two. More than half, 52%, showed up in only one run.
So being named once tells you very little. If a customer or a rep tells you ChatGPT mentioned your shop, ask the same question again, logged out. Then ask it a third time. A shop that comes back every time is in a different spot from one that showed up once.

Google and ChatGPT point buyers to different sites

Only 6% of the sites ChatGPT cited for a question were also in Google's top 10 for the same question.

Shops' own sites

ChatGPT (90 answers)

Cited in 88% of answers

Google (30 searches)

44% of the top three regular results

Map listings

ChatGPT (90 answers)

Cited in 32% of answers

Google (30 searches)

A local map pack on 24 of 30 searches

Industry associations or networks

ChatGPT (90 answers)

Cited in 39% of answers

Google (30 searches)

3% of the top three regular results

Google also showed an AI Overview on 17 of the 30 searches. Being found on Google and getting named by ChatGPT are two separate checks.
Asked from the US, the way a Detroit-side buyer would, the pattern held. On 10 of the questions, own sites were cited in 26 of 30 answers (87%) and associations in 50%. Two of the 30 drifted: one to a Windsor in California, and one asked which Windsor we meant.

What to check about your shop this week

None of this needs a tool.
  1. Ask ChatGPT the question your buyers ask, in their words, while logged out. Include the process, the spec and the place. Ask it three times and write down whether your shop is named every time, once or not at all.
  2. Read your home page the way a buyer would. Does it say plainly which processes you run, which certifications you hold, what you specialize in and how far you are from the major shipping hubs?
  3. Check where that information lives. If your certifications or equipment list are only in a PDF, put them on a regular page too.
  4. Look up your shop in the industry networks and directories your buyers' answers point to. Is it listed, and does the listing say the same thing your site says?
  5. Search your shop's name on Google. Is the first result you, or a company with a similar name? Does Google's business panel link to your site?

Where to start with your shop

What gets a shop named changes from one industry to the next, and with what the buyer is trying to do. That's why we track it across industries.
If you run a job shop, a tool and die or mold shop, a coating line or a contract manufacturing plant, our manufacturing page is the place to start. If you'd rather we take a look, the AI Search Game Plan is a one-time review of how AI assistants describe your shop and what to fix first.

How we ran this study

  • What we asked: 30 questions a plant buyer might ask when looking for a supplier near Windsor-Essex and the Detroit border, in five groups: finding a process nearby, a job with a spec, building a shortlist, sourcing across the border, and one supplier for more than one job.
  • How: each question three times in ChatGPT through DataForSEO's ChatGPT scraper, from a Canada location (country level, not Windsor-Essex itself), logged out. That's 90 answers. Each question also ran once on Google from Canada, with AI Overviews loaded.
  • The US check: 10 of the same questions, three times each in ChatGPT from a US location. That's 30 more answers.
  • The name check: the 40 shops ChatGPT named most often, among those named in at least two of three runs of the same question, each searched once on Google Canada by name.
  • The schema check: 30 of the pages ChatGPT cited most often (7 network or association pages, 9 directory pages and 14 shop pages), each fetched and read for structured data. 22 could be read; the rest blocked an automated fetch. Small, and it shows what the pages carry, not why they got cited.
  • How we sorted it: each cited site tagged by hand by type. Directories means business directories and supplier marketplaces, such as Yellow Pages and Thomasnet. Industry associations and networks, map listings and social sites are counted separately. Each named company was coded by type from the answer text.
  • When: October 20, 2026.
  • Limits: a sample on one day, not a census. We have no RFQ or sales data, so this shows where ChatGPT and Google point a buyer, not who got the work.
  • We plan to run the same questions again and report what changed.

The questions we asked

The 30 questions, as we asked them:
  1. Powder coating companies in Windsor Ontario
  2. CNC machine shops in Windsor Ontario
  3. Tool and die shops in Windsor Essex
  4. Injection mold makers in Windsor Ontario
  5. Metal stamping companies near Windsor Ontario
  6. Metal fabrication shops in Windsor Ontario
  7. Laser cutting services in Windsor Ontario
  8. Robotic welding companies in Windsor Essex
  9. Plastic injection molding companies in Windsor Ontario
  10. Automation and robotics integrators in Windsor Ontario
  11. I need a powder coating supplier for OEM aluminum panels, high volume, ideally one who can also do fabrication, within about an hour of Windsor
  12. IATF 16949 certified machine shop in Windsor Ontario
  13. Who can do fast prototype CNC parts near Windsor or Detroit
  14. Low volume sheet metal fabrication on the Canadian side near Detroit
  15. Who can build large injection molds for automotive parts in Ontario
  16. ISO 13485 contract manufacturer in Southwestern Ontario for medical device parts
  17. Best tool and die shops in Windsor Essex
  18. Top mold makers in Canada
  19. Contract manufacturers in Southwestern Ontario
  20. Which machine shops in Windsor do aerospace work
  21. List of automotive suppliers in Windsor Ontario
  22. Should I source tooling from Canada or the US for a Michigan plant
  23. Suppliers in Windsor that ship parts to Detroit plants
  24. CUSMA compliant automotive parts suppliers in Ontario
  25. How do I find a reliable contract manufacturer in Ontario
  26. How do I vet a new machining supplier before sending an RFQ
  27. One supplier for metal fabrication and powder coating near Windsor
  28. Machine shop near Windsor with capacity for large parts
  29. Who does both tool design and tool build in Windsor
  30. Short run metal stamping supplier in Ontario
The US check asked questions 1, 2, 3, 4, 11, 13, 17, 23, 25 and 27 with the same wording, three times each from a US location.
More from here: how we work in Windsor-Essex and across Canada. Another field study: what ChatGPT named on Maui.

Common questions

What should marketing for manufacturers focus on now that buyers ask ChatGPT for suppliers?

In our Windsor-Essex runs, ChatGPT cited a shop's own website in 79 of 90 answers, and 64% of those citations went to home pages. So the shop's own site, and the home page first, is the place to check. It should say plainly which processes the shop runs, which certifications it holds and what it specializes in.

Does ChatGPT cite certification pages and PDFs on a shop's website?

Rarely, in our runs. Of 408 citations to shops' own sites, 4 were quality or certification pages (3 of them the same page), and PDFs were 13 of all 704 citations. The study can't tell whether those pages were missing or just weren't read. Putting certifications and capabilities on a regular page as well as in a PDF is a simple step.

Why does ChatGPT cite industry associations and business directories when buyers ask for suppliers?

We measured what got cited, not why. An industry association or network showed up in 35 of 90 answers (39%) and a business directory in 29 (32%). My read is that these sources hold curated data organized around what buyers check, in a standard layout, and get used because answers are built from many sources at once. On the cited pages we could read, 6 of 10 network and directory pages showed a structured list of businesses. Of 12 shop pages, 1 described the business itself in schema.

What happens when a buyer searches my shop's name on Google?

We searched 40 of the shops ChatGPT named most, by name, on Google Canada. For 8 of them, the results were mostly about a different company with a similar name. The shop's own site came first for 18, and Google's business panel showed up for 33, with reviews on 30.

If ChatGPT named my shop once, will it keep naming it?

Not necessarily. We asked each question three times. Of the companies named for a question, 21% came back in all three runs and 52% showed up only once. Ask the same question a few times, logged out, before reading much into one answer.

Does this study show which shops won the work?

No. We measured what ChatGPT cited and named, and what Google showed, on October 20, 2026, logged out, from Canada with a 10-question check from the US. We had no RFQ or sales data, so we can't say which shops got asked to quote or got the job.

Sources

  • Plate Lunch Collective. Windsor-Essex manufacturing field study: 30 supplier questions with three ChatGPT runs each from Canada, Google results with AI Overviews for the same questions, a 10-question check from the US, and a 40-shop name search on Google Canada. October 20, 2026. The question list is in this post; summary data on request.
Share this article
Add as a preferred source on Google

Topics

Ready to appear in AI search?

Plate Lunch Collective is an AI SEO agency that gets brands named when buyers ask, in search, on social, and in AI. If you have questions about where you stand, we're easy to reach.

Work with us

Hayden Bond

Hayden Bond has been doing SEO since 2004. He founded Plate Lunch Collective in Aiea, helping brands get cited by AI platforms rather than just ranked by Google.

Stay current on AI search

New research and analysis, direct to your inbox.

Related articles

Explore our AI Search Visibility services

Previous

Vacation Rental Marketing on Maui: What ChatGPT Names and What Google Shows Next