Guide · Method
How manufacturers say what they do: the method, and how to check your own site
This guide sets out how we measure how much of what a plant can do its website actually says, and how to run the same check on your own site in an afternoon. We are running it as a study across shop types. We'll publish the first results here when the study is done, with the sample, the method and the date.
Who wrote this
Industrial Marketing wrote this method and is running the study. We market industrial companies, so we have a view on the answer. That is why the method is published before any result, and why every plant in the study takes part with its permission. Checked October 2026.
Why
Why measure how manufacturing websites say what plants do
Buyers can only shortlist what they can find and check. An engineer searching for a process, a buyer screening suppliers and an AI assistant answering 'who can make this' all work from what a website says. A capability that isn't written down doesn't exist to any of them.
We see the gap on on-site visits: processes run every day that never reach the site, secondary ops nobody mentions, certifications shown as a logo with no scope. Owners usually know their site is behind. Few know by how much, or which kind of capability goes missing most.
The study will put a measured, dated number on that gap, by shop type. It will give an owner a fair comparison with plants like theirs, and give us a figure to cite in place of a rule of thumb. You can't sell what you don't say. This is our attempt to count what goes unsaid.
Method
Method
Five steps, the same for every plant in the study.
-
Step 1, OP 10
Sample
Machine shops, fabricators, coatings and finishing, connector makers, medical suppliers and equipment makers. Each one agrees to take part.
-
Step 2, OP 20
What we counted
Processes, materials, envelopes, secondary ops and certifications: the facts a buyer screens on. Each one becomes a line in the plant's capability list.
-
Step 3, OP 30
How we check the floor
Against each plant's own documents and interviews, with permission. A capability counts only when the plant has confirmed it.
-
Step 4, OP 40
How we read the site
Each confirmed capability is marked said, partly said or not said, from the public website only, read the way a buyer reads it, on a stated date.
-
Step 5, OP 50
Limits
What the sample can and can't tell you, set out in full below.
Scoring
What counts as said
Every confirmed capability gets one of three marks. The share marked said is the main figure, and partly said is reported beside it.
| Mark | What it means | Example |
|---|---|---|
| Said | Named in words on a page a buyer can reach, with the detail a buyer screens on | Swiss turning to a stated bar diameter in titanium and 316L, on a process page |
| Partly said | Named without the detail, or only inside a PDF, a photo or a logo | 'Precision machining' beside an uncaptioned photo of a Swiss lathe |
| Not said | Confirmed on the floor, missing from the website | In-house passivation that appears nowhere on the site |
What we count
The five things we count
These are the facts a buyer checks before sending a drawing. Each one is counted separately, so the study can show which kind of capability goes unsaid most.
-
Processes
The operations the plant runs: machining, forming, joining, coating, molding, assembly and test.
-
Materials
Alloys, grades, resins and substrates, named the way an engineer specifies them.
-
Envelopes
Part size, weight and tolerance limits: the numbers a buyer uses to check fit before calling.
-
Secondary ops
Work after the main process: finishing, inserts, assembly, marking and test.
-
Certifications
Quality registrations and approvals, counted as said only when the scope is stated too.
By shop type
What each shop type's buyer screens on
The five counts are the same for every plant. Which lines matter most changes by shop type, so each result is read against what that buyer checks first.
| Shop type | What its buyer screens on first |
|---|---|
| Machine shops | Machines and their envelopes, materials, tolerances, quantities and secondary ops |
| Metal fabricators | Materials and thicknesses, bed and brake sizes, welding processes and codes, finishing |
| Coatings and finishing | Process and spec names, substrates, thickness, part size limits and approvals |
| Medical device suppliers | ISO 13485 scope, materials, cleanroom work, validation and traceability |
| Aerospace and defense suppliers | AS9100 scope, special process approvals, ITAR registration and first article inspection |
| Equipment makers | Applications, throughput, footprint, options, service and parts |
Data
Get the first results
We'll publish the first results here when the study is done. They also go to The Supply Side, our monthly newsletter. This page stays open to read.
How to use it
How to use this method
Owners: run the check below on your own site, using what your shop type's buyer screens on first. The lines you mark not said are the pages to write first.
Marketers and agencies: use the five things we count as the questions for your next interview with an engineer.
Manufacturing associations: run the check with a few member shops to see where members' websites fall short as a group, which helps when you plan a member directory or a workshop.
Your own site
Run the same check on your own site
You can run a rough version of this check on your own website in an afternoon.
-
Step 1, OP 10
List what you run
Walk the floor with a notepad: every process, material, envelope, secondary op and certification. Ask estimating what they quoted last month that isn't on the list.
-
Step 2, OP 20
Find each line on your site
For each line, find the page a buyer would land on. Note whether it names the item in words, with the numbers.
-
Step 3, OP 30
Mark each line
Said, partly said or not said, using the table above.
-
Step 4, OP 40
Count
Divide the lines marked said by the total. That is your share. The lines marked not said are your next pages.
-
Step 5, OP 50
Ask an AI assistant
Ask who can do three of your not-said items. See whether you are named, and which pages it cites.
Limits
What this method can't tell you
A study made of companies that agree to take part may lean toward plants already thinking about their websites. A sample can show a pattern by shop type. It can't stand in for every plant in a market. A capability counts only if the plant confirmed it, so anything the plant forgot to mention is missing from both sides of the count.
Each website is read on a stated date, and sites change. Published results will carry their date, and we name no company in the sample.
FAQ
Questions
What share of their capabilities do manufacturers' websites state?
We'll publish the measured share here when the study is done, with its sample and date. Until then, our rule of thumb from working with manufacturers is about 40%. It was never measured. One measured case, in March 2026, came out at about 52% of confirmed capabilities listed.
Which companies are in the sample?
We won't name them. Each takes part with permission, and published results will identify plants only by shop type and market. Examples elsewhere on our site use invented companies with .test web addresses.
Can I cite this page?
Yes, for the method. There is no measured figure on it to cite yet. Link to this page so readers can check the method behind any figure you quote.
See your own share, line by line
Start with the free Buyer's-Eye Audit: write the RFQ you most want to win, and we'll show you where a buyer would shortlist you or drop you. The Buyers Blueprint, our paid first step, maps what your plant does against what your site says, for every capability a buyer screens on.