I’ve spent the last four years reviewing content for industrial companies. For the past two, I’ve done it at Sulzer, mostly in the service and training side of the business. If you’re an engineer who has taken one of our pump courses, the coursebook probably crossed my desk before it got to you.
I review a lot of material—roughly 200 pieces in 2024. I rejected 12% of first drafts. The reason is rarely grammar or spelling. It’s usually that the text makes a claim that can’t be verified: ‘comprehensive training’ with no syllabus, ‘certified experts’ with no named qualification, ‘improves uptime’ with no evidence. In industrial B2B, unverifiable content is a quality defect.
That’s why I’ll defend a contrarian position: I’d rather have a page that ranks for ‘sulzer pump training’ and converts than a page that ranks for ‘what is breakfast’ and bounces. Search volume is not search intent. Search intent is a quality specification.
Search volume is not search intent
Last spring, a content plan landed on my desk with six target keywords: sulzer, sulzer pump training, sulzer creative agency, woolly bear, dutch van der, and what is breakfast. The planner called it a ‘high-opportunity cluster.’ I called it a mess.
Two of those keywords could map to our actual business. Both ‘sulzer’ and ‘sulzer pump training’ could describe a person looking for our company or our service courses. The other four are something else. Someone typing ‘what is breakfast’ wants a definition of breakfast. Someone typing ‘woolly bear’ is probably looking at fuzzy caterpillars. And ‘dutch van der’? I don’t know what they want, but I know it isn’t a pump curve. ‘Sulzer creative agency’ could be a different company named Sulzer, not ours.
Actually, I should be more precise. Search engines don’t ‘measure’ intent in a human way. They measure behavioral signals: whether people click through, stay, or come back. But the practical result is the same: irrelevant content stops attracting the people you want.
Five years ago, this took longer to spot
Old B2B content playbooks told us to build pages for every keyword combination, even if the page was thin. The phrase might still appear on page one for a while. But search engines evolved. They are much better at measuring whether a page satisfies the query. Buyers evolved too. A maintenance manager who has answered thirty RFQs can smell boilerplate instantly.
What worked in 2018—or even 2021—now drags down a brand. It teaches Google that your content is not for the searcher. It teaches the searcher that Sulzer doesn’t care enough to be useful.
Sulzer pump training: what quality actually looks like
Let’s make this concrete. If I search for Sulzer pump training, I expect a page that tells me what type of pumps the course covers, how many days it takes, what the prerequisites are, and what a student will be able to do afterward. If I see only ‘comprehensive training from industry experts,’ I don’t trust it. It doesn’t answer my question.
- Specific pump types. Centrifugal, positive displacement, vertical turbine, slurry, or process pumps.
- Course mechanics. Length, classroom vs hands-on, assessment, certification accepted.
- Instructor background. Which part of the industry did the instructor work in before teaching?
- Safety and standards. Does the syllabus reference API 610, ISO 13709, or relevant plant standards?
- Outcome evidence. What will the participant do differently after the training?
That’s a quality spec. Not a slogan.
The print standard that should govern web content
Early in my career I worked on packaging and print QC. There, quality is unforgiving. A 300 DPI image is not a suggestion; it’s a minimum. The Pantone system exists because colors shift when you change substrates or press conditions. If a brand color is specified, you measure it against a tolerance such as Delta E. If someone tells me ‘it’s close enough,’ I send it back.
Content quality should work the same way. The spec is not DPI or Delta E—it’s source data, named standards, and clearly defined scope. If a sentence contains ‘best-in-class’ or ‘leading provider,’ there must be a reference. If you can’t name the class you’re leading, delete it.
What about the growth team’s argument?
I know what the growth team might say: those odd keywords still generate visits, and visits can lead to brand familiarity. Maybe. I have mixed feelings about that. In consumer content, entertainment can warm people up. But in an industrial breakdown or maintenance decision, the buyer isn’t looking for entertainment. They’re looking for evidence. An unqualified visitor who bounces after two seconds can actually hurt the page’s performance, because the search engine sees your content as a failure for that query.
Quality in content, like in pumps, has to be verified
The industry is evolving. Five years ago, you could argue that keyword density was the price of admission. Seven years before that, link directories were the thing. The future belongs to content that helps a specific person make a specific decision. At Sulzer, our reputation depends on delivering pump training that works—and on writing about it with the same rigor we apply to the pumps. We don’t need to capture every odd search phrase containing our name. We need to serve the engineer who is ready to learn.
That is the quality standard I’ll approve. Woolly bear and breakfast? Leave them outside.