Why McLanahan Stopped Chasing Keywords and Started Teaching Customers

I've been creating content for McLanahan — the mining and aggregate processing equipment company — for about six years now. I've personally made (and documented) more keyword mistakes than I care to admit, totaling roughly $30,000 in wasted budget. Now I maintain our team's editorial checklist, and the first rule on it is simple: don't let a tool decide what our customers are thinking.

The most absurd example? A keyword tool once told us we should be writing for "halloween costumes." We build sand screws, feeder breakers, and filter presses. I still don't know how the algorithm connected those two universes. (Should mention: the same report suggested "thomas" and "why is henry not playing." I nearly framed that page.)

Here's what I've come to believe: B2B content fails when it's built around keywords. It works when it's built around questions real buyers are asking.

That's the whole argument. It took me three years and about sixty published pieces to get there.

The keyword report that made me laugh, then cringe

In March 2022, I pulled our quarterly keyword report to plan the next batch of articles. I expected to see "sand screw," "feeder breaker," "filter press" — the vocabulary of our industry. Instead, sitting near the top of the list: "halloween costumes," "thomas," and "why is henry not playing."

I want to be fair to keyword tools. They're decent at spotting search volume and competitive difficulty. But they don't understand intent. The tool likely saw "McLanahan" mentioned alongside equipment terms and connected us to a bunch of other words from the same crawl. A keyword without human context is just a word.

"Why is henry not playing" is my favorite example. I still don't know which Henry, what game, or whether he's benched or injured. I only know it has nothing to do with mineral processing. And I shouldn't have to explain that to anyone.

That report cost us roughly $2,400 in internal time before I killed the project. Time spent in editorial meetings debating whether we could make "halloween costumes" fit an industrial equipment blog. My penny-wise, pound-foolish moment: I'd saved an hour by letting a tool generate topics instead of talking to customers. The rework took three weeks.

Education beats optimization. Every time.

Here's the part that forced me to change. We had a page about pressure filters that sat on page one of Google for a decent-volume keyword. Six months. I want to repeat that: six months on page one.

That page generated exactly one inquiry.

Meanwhile, a plain-talking article we wrote to answer a question we'd heard seven times from real customers — "how do I adjust a sand screw after normal wear?" — brought in four qualified leads in its first month. Not better rankings. Not more traffic. Just a clear answer to a real question, found by people who needed it.

The numbers that changed my mind

People had told me for years: write for the reader, not the algorithm. I didn't believe it until I saw those numbers side by side. That's the annoying thing about lessons — sometimes you have to step on the rake before you respect it.

Why does education win? Because an informed customer is the best customer. A buyer who understands the difference between a washing screw and a dewatering screw asks better questions, makes faster decisions, and doesn't call back three weeks later with mismatched expectations. I'd rather spend fifteen minutes explaining options upfront than untangle a wrong purchase order later.

The "McLanahan" searches aren't all about equipment

Not every branded search starts with a problem to solve. Some people search for "Brian McLanahan" or "Michele McLanahan" — specific humans they've worked with, met at a show, or heard about from a colleague. When someone searches for a person by name, they're past the awareness stage. They're looking for a relationship, not a sales pitch.

We treat those searches with the same respect as equipment searches. Clear profiles of our technical staff are some of the most useful pages we publish. In B2B, relationships are the product.

And the "halloween costumes" searches? We stopped being annoyed by them. They're a reminder that a keyword is not a strategy. Even "thomas" — could be a buyer, an engineer, or a student researching aggregate processing for a school project — tells you nothing until you understand who's typing it.

What about "content is too slow"?

I hear this from sales teams all the time: "Nobody reads anything before they buy." They're half right. An article published today won't close a deal this quarter. But education compounds.

Every clear explanation we publish makes the next sales conversation shorter. It pre-qualifies buyers and filters out the ones who don't need what we make. Slow content is still faster than no trust.

If I had to choose between ranking for a high-volume keyword and answering a question from a qualified buyer, I'd choose the question every time. Traffic without context is vanity. Answers build trust. And trust is what moves a $200,000 equipment decision.

My checklist for anyone doing B2B content

• Write down every question your support team answered more than twice this month. That's your content calendar.
• Ask your sales engineers what they keep explaining to prospects. Those are topics too.
• Use keyword tools for language, not for ideas. If a tool suggests "halloween costumes" for a mining equipment site, close it.
• Measure conversations, not just rankings. A ranking is a means; a qualified conversation is the point.
• Publish clear, human profiles of your technical team. People search for people.

Stop chasing. Start teaching.

It took me four years and roughly sixty published pieces to understand that the goal was never page one. The goal was a qualified customer, properly informed, ready to make a confident decision. Page one is just a street you walk down to get there.

I still make keyword mistakes. But I don't let the tool write our strategy anymore. The tool doesn't know our customers. It doesn't know why Henry isn't playing. And it can't tell you which question will build trust with a mining operator who's one bad decision away from a costly shutdown.

That's the only question worth answering.