Case study: how Clean screened 6,214 US manufacturers and cut three of five message angles

The seller, the ICP and five message angles

The customer sells a growth product to mid-size US manufacturers. We're keeping the company and the product out of this write-up. Their ICP was specific: US manufacturers with $20M or more in revenue, 45 to 250 employees, founded in 2015 or earlier. That's an established company with real revenue, still small enough that the owner or GM is often the one who signs.

They came with five approved message angles. Each angle was a reason to write to one specific company, and each rested on a single fact about that company. When the fact is true, the note tells the reader something they recognize about their own business. When it's false, the note makes a claim they can disprove before they finish reading it.

So before the list got divided across five angles, Clean checked how many companies each fact was actually true for.

How many of 6,214 manufacturers had a named decision-maker

Clean worked through 6,214 US manufacturers that fit the ICP and matched a named decision-maker at 4,671 of them, about 75%, measured on September 8, 2026. That left 1,543 companies without a named person. Plan volume around the 4,671, not the 6,214.

The size floor is doing real work here. The National Association of Manufacturers counts more than 239,000 US manufacturers, and 74% of manufacturing firms have fewer than 20 employees. A generic manufacturers list is mostly small shops. A 45-employee floor drops all of them, and the revenue and founding-year cutoffs narrow what's left again.

Who counts as the decision-maker changes with size. As a typical pattern, the owner of a 50-person shop often signs personally, and around 200 employees a president or GM signs after the controller checks the payback. In this band, the name worth finding is usually the owner, president or GM, with the controller close behind. For the full approval path, see how manufacturing plants buy.

Testing each message angle against every company first

Before anything went out, Clean tested each of the five angles against every company on the list. One question per angle: at how many of these companies is the underlying fact true, and would it survive the reader checking it?

That's a harder bar than whether a personalized first line can be written. A first line can always be written. But a manufacturer knows its own order minimums, buildings and customers better than any seller does, so a wrong fact gets caught fast.

Why the minimum order quantity angle was cut

This angle needed the company to publish a minimum order quantity. Of 6,214 companies, 46 publish one at all. Only 19 of those give an actual number. The other 27 mention a minimum without saying what it is.

Our read on why: plenty of mid-size manufacturers, contract shops especially, price each job from the part, the material and the setup, so any minimum shows up in the quote rather than on the website. Whatever the reason, the ceiling was 19 companies out of 6,214. That's a handful of hand-written notes at most. Pushing it to volume would have meant telling companies about a number they never published.

Why the expansion angle found zero companies

This angle told a company that its own website hadn't caught up with its growth. For it to work, two things had to be true: the company really had expanded, and its site still made a claim that the expansion had made out of date.

About 160 companies used launch language that reads like growth at a glance. Every candidate fell apart on a closer look. Each one turned out to be an equipment purchase, an acquired building or a customer's facility, and none was an expansion the site had failed to reflect. The angle shipped to zero companies.

That lesson carries to anyone selling into plants. Growth language on a website and new capacity at a plant are different things. A new machine, a bought building and a customer's facility are three separate events, and each calls for a different note, or none at all. Finding real capacity early is its own discipline, covered in new factories and plant expansions.

The angle that rarely checked out, and the two that held

The third cut depended on a specific claim about each company. It turned up candidates, and almost none of them held up on verification. Run without checking, an angle like that puts a false claim in nearly every note it produces, and the reader is the one who catches it.

Two angles held. Each rested on a fact that stood up company by company, and those two carried the volume. The campaign went ahead on two angles instead of five, with the split between them following where the facts were true instead of an even share. We're keeping the surviving two private, since they're the seller's playbook.

How to test your own message angles before a campaign

Clean would rather hand a seller fewer reasons to reach out, all of them true, than fill a quota with claims a manufacturer can disprove. A wrong fact about someone's own business is a fast way to lose that reader. Why generic notes stall with this audience is covered in cold email to plant managers.

You can run the same test on your own angles before the next campaign:

• Write down the one fact each angle depends on, as a sentence the prospect could confirm or deny.
• Count the companies on your list where that fact is true before anyone writes a line.
• Study the near misses. For the expansion angle they were equipment purchases, acquired buildings and customers' facilities, and each would have become a false claim.
• Set a floor. An angle that holds at only a few dozen companies gets cut, or saved for a few hand-written notes.
• Put the volume where the facts hold, even if the split ends up lopsided.

What Clean does for companies that sell to plants

The rule that cut three of five angles here runs through all of Clean's manufacturing work. No reason to reach out ships without its date and the evidence behind it, and each one lists the other explanations that could account for it and the finding that would disprove it. Where Clean can't confirm something, the answer is unknown, not a guess. The walkthrough is in how Clean works.

For sellers into plants, those reasons come from the 14 buying moments Clean groups a plant's life into, researched plant by plant rather than only at headquarters. That's the gap between a list of manufacturers and a buying moment you can act on. If you sell CMMS, MES, quality or automation into plants, the manufacturing overview shows how this works for your category. Book a demo; before the call ends, Clean will have a live list of plants built for your product.