Why Your Keyword Strategy Is Ruining Your Content Marketing (A Case Study in Bad Data)

The day the report landed on my desk

It started with an email. Subject line: "SEO keyword analysis for Q2." The report had 10 keywords. Someone had spent real money on it. And they were, basically, a mess.

I'm a procurement manager. I spend my days looking at spreadsheets of vendor costs, inventory levels, and shipping quotes. But over the past six years of tracking every invoice and analyzing $180,000 in cumulative spending, I've learned one universal truth: garbage data produces garbage results.

This was garbage data.

The keywords were: fuller, robert fuller tv shows, fuller house danny, millennium lego, eddie kids, difference between and hawk.

I stared at my screen. I blinked. Then I laughed out loud.

Our company is in the energy and mineral equipment industry. We sell parts for crushers. We optimize grinding mill circuits. We don't do Lego. We definitely don't do TV shows.

But instead of just rejecting the report, I decided to dig in. Because honestly, this is the kind of thing that happens way more often than people admit. And it's costing real money.

The audit: What was actually in the data?

I pulled up our cost tracking system. We'd spent $4,200 on that keyword analysis—part of a larger marketing package. The vendor had promised "comprehensive market insights." What they delivered was a glorified Google search dump.

Let me break down the breakdown:

• "fuller" — One keyword. Our brand name. But zero context. No modifiers like "fuller mining equipment" or "fuller crusher parts." Just the word. A brand search that tells you nothing except that people search for the brand. Not useful for strategy.
• "robert fuller tv shows" — An actor. From the 1950s. Not our target audience. I can say with confidence that no one looking at crusher gearboxes is also looking for classic TV Westerns.
• "fuller house danny" — A sitcom character. Zero overlap with mineral processing.
• "millennium lego" — A toy set. For children. Not for industrial procurement managers.
• "eddie kids" — Could be a clothing brand. Could be a person. No business intent.
• "difference between and hawk" — This one broke me. It's an incomplete query. It doesn't even have a subject. Someone typed "difference between blank and hawk" and walked away.

Of the 10 keywords, I could justify exactly zero for our B2B content strategy. Maybe one—maybe—if you squinted really hard. The noise-to-signal ratio was 90%. That's not analysis. That's noise.

I went back and forth between confronting the vendor and just quietly fixing it myself for two weeks. On one hand, they should know they delivered a bad product. On the other, I didn't want to burn the relationship if they could fix it. The established vendor vs. the headache of starting over — this is the classic procurement dilemma.

The real surprise

But then something unexpected happened. When I looked closer at our actual search traffic data, I found a pattern I hadn't expected.

Our best performing content wasn't for the sexy keywords nobody searched for. It was for the boring, specific, functional queries like "conveyor belt tension calculation" or "vibrating screen deck replacement interval." These had search volumes of 20 to 100 per month. Tiny numbers. But the conversion rate? 12%. Compared to our generic brand keyword conversion rate of 0.8%.

The surprise wasn't the price difference. It was how much hidden value came with the "boring" option—genuine help for people who needed answers. Their searches were specific. Their intent was clear. They weren't browsing. They were looking for a solution to a problem they were already dealing with.

I knew I should push back on the keyword vendor immediately. But I also knew that arguing with a vendor over a $4,200 report is a time sink. So instead, I pivoted our internal strategy.

Skipped the final review meeting where we would have approved the next phase of that vendor's work because we were rushing and "it's basically the same process as last time." It wasn't. Not using that report saved us from another $4,200 in wasted spending.

The bottom line

There's something satisfying about a perfectly executed pivot. After all the stress of seeing a bad report, the frustration of realizing someone took short-cuts, and the temptation to just let things slide—finally, making a decision that actually improves outcomes. That's the payoff.

In Q2 2024, when we switched our content strategy to focus entirely on customer education around specific technical pain points, our inbound qualified leads went up by 17%. Our total content spend went down by 22%. An informed customer asks better questions and makes faster decisions. Give them useful answers, and they'll find you.

So here's what I learned from this whole ordeal:

1. If a keyword can't pass the "would a real buyer type this?" test, it's noise. TV show characters and toy brands don't buy mining equipment. It sounds obvious, but it's amazing how often people forget this when they're staring at a spreadsheet.
2. Ignore the vanity metrics. High search volume for irrelevant terms is just expensive noise. Low-volume, high-intent terms are where the real ROI lives.
3. Garbage data costs more than just the report fee. It costs the time you spend auditing it. It costs the opportunity of not doing the right thing. And it costs the trust you place in a process that should work but doesn't.

I'd rather spend 10 minutes explaining to a marketing manager why a keyword list is useless than spend the next quarter writing articles about the difference between a hawk and a dove for an audience that doesn't care. That's not content marketing. That's just noise with a budget.

Sometimes the most valuable thing you can say is: this data is wrong. Let's go back to step one. Because step one—understanding what your customer actually needs—is the only one that matters.