Popular keywords are a trap. I know that sounds obvious now. I ignored it anyway, and in Q3 2024 it cost us $3,200 and two months of pointless clicks.
For context, I handle marketing for Falcon, a B2B supplier of components for energy and mining equipment. I've been doing B2B marketing for seven years, and I keep a running list of my own mistakes. This one reached the top of the list pretty fast.
Our name is easy to remember, which sounds like an advantage. It's not—not in search, anyway. "Falcon" is one of the most overloaded words on the internet. It's a cybersecurity product (CrowdStrike Falcon), a spaceship in Star Wars, the title of a movie (The Bone Collector), a local service business called Eddie's, and I once saw a search for "white vs magic" in our keyword data that still doesn't make sense to me. If you sell industrial parts, you probably don't want to be on that page. But I paid to put us there.
Look, I'm not saying brand keywords are always useless. If you're a consumer brand with a unique name, they might work well. But for a small B2B manufacturer whose name collides with pop culture, relying on "falcon" was the wrong bet. The lesson was expensive, and it changed how I think about search intent. Let me show you what happened.
What the Search Term Report Actually Looked Like
After the first month of that campaign, I pulled the search term report expecting to see "falcon mining," "falcon components," maybe "falcon parts." Instead, the clicks came from things like:
"arctic wolf crowdstrike falcon"
"star wars falcon"
"bone collector"
"eddie near me"
"white vs magic"
and about four hundred variations of "falcon."
Here's the thing: none of these people were looking for mining equipment. "Arctic wolf crowdstrike falcon" is someone trying to install a security tool, not a conveyor system. "Star wars falcon" is probably a fan. "Bone collector" could be someone looking for a movie. "Eddie near me" might be someone looking for a barber. We were paying for those clicks.
Total after the quarter: 1,214 clicks. Qualified inquiries: zero. That was a fairly uncomfortable meeting, and I've had a few.
What Happened When We Tested the "Boring" Keywords
Around the same time, we ran a small parallel test on specific product terms like "mining equipment replacement components" and "custom industrial coils for energy projects." The search volume was kind of embarrassing—maybe 90 to 150 searches per month. I told my manager it wasn't worth the effort. I was wrong.
When I compared the two campaigns side by side, the contrast was obvious. The broad campaign brought 1,214 clicks and no leads. The narrow campaign brought 37 clicks and four requests for quote. One of those requests turned into a $28,000 order. There's something satisfying about a campaign that gets 37 clicks instead of 1,214 and still produces a meeting with a real buyer. After the stress of the first quarter, it felt like proof that we had been speaking to the wrong room the whole time.
A lot of the advice I read about keyword research talks about search volume. But volume is not intent. People search "falcon" because they want a movie, a game, a bird, or a security update. They search "industrial coils for mining equipment" because they have a problem that needs a solution. I was choosing the wrong metric because it looked better on the dashboard.
Why "Falcon" Keeps Failing as a Search Term for a Manufacturer
Think about how an engineer sources a part. She doesn't type a vendor's name into Google and hope the right company appears. She types the part number, the material spec, the failure mode. She searches "high-torque spring for mining conveyor" or "explosion-proof enclosure 480V." Then she looks at the results, checks the datasheet, and only then does she visit the brand site.
Our name meant nothing at that stage. The only way to show up for her is to have content that matches the spec. A brand campaign around the word "falcon" is a distraction. It feels visible, but it's visible to the wrong people.
But What About Brand Awareness?
You could argue that being in front of 1,214 people is worth something. Under normal circumstances, I'd agree. But with a $3,200 budget over three months, we needed real conversations, not impressions. Awareness is a nice idea until the invoice arrives.
There's also a question of honesty in advertising. Per FTC guidelines (ftc.gov), marketing claims should be truthful and not misleading. Showing an ad for "falcon" to someone searching "white vs magic" isn't illegal, but it's misleading in the sense that the searcher has no idea why they're seeing it. It wastes their time and our money.
If we wanted awareness, the smarter place was trade shows and LinkedIn, where the audience already knows they work in energy and mining. Paid search is for capturing demand, not manufacturing it out of vague keywords.
What I Do Differently Now
I didn't delete the word "falcon" entirely. That would be the easy, absolute solution—and honestly, not always the right one. Instead, I moved it to a phrase-matched campaign with negative keywords like "crowdstrike," "star wars," "millennium," "bone collector," and local service terms. That small fix reduced wasted spend by about 90% in the first month.
I also stopped approving campaigns based on keyword lists alone. Every keyword now has to pass a simple test: if a stranger read this search term out loud, would they know what the searcher wanted? "Bone collector"—no. "Mining equipment replacement components"—yes. That's the difference between a traffic strategy and a lead strategy.
The Version I Wish I'd Started With
There's no perfect keyword. There's only the keyword that matches a real buyer's question. For Falcon, that meant abandoning the word "falcon" almost entirely and building pages around product categories, materials, and industry names. It felt like giving up. It was actually the first smart thing we did.
The $3,200 is gone, but the lesson stuck. Don't chase volume if the volume doesn't buy what you build. And if you ever see "white vs magic" on your search term report, pause the campaign before your boss sees it.