Understand the Human · Search & Advertising
The basic truth
You named your product. Then you bought your product’s name.
But the person who needs what you sell usually doesn’t know it’s called that. They know what’s going wrong on their floor, what they’re trying to make, what material is giving them trouble, or what the last machine couldn’t do. They search the problem, because the problem is the only part they’re sure about.
This is the step before the promise. Before you can send a click somewhere honest, you have to know what the person was actually asking — and that isn’t a keyword question. It’s a question about a human being who has a job to do and a decision they’re nervous about.
What to do
Start with the vocabulary gap. Write down how you describe what you sell. Then write down how a customer described it to your sales team last week. If those two lists match, one of them is wrong — and it isn’t the customer’s.
Industrial buyers search by problem, application, material, industry, comparison, and specification — by what they’re trying to accomplish, not by what you’ve decided to call it.
Then stop looking for “the buyer.” There isn’t one. A capital purchase moves through several people who search differently for the same machine, and a keyword list built for one of them is invisible to the rest:
| Who’s searching | What they type | What they’re really asking |
|---|---|---|
| The engineer | Specifications, tolerances, materials, comparisons | Will this actually do the job? |
| The operator | “How do you…”, setup, training, maintenance | Can we run it without it becoming my problem? |
| The owner / CFO | Cost, ROI, lead time, financing, “worth it” | What happens to me if this is the wrong call? |
| Purchasing | Brands, models, terms, availability, alternatives | Can I defend this choice to the people above me? |
They’re not competing searches — they’re one decision, seen from four seats. (This is Facet 1, Map the Buying Decision, doing its work.)
Two places to find the real language, both free and both underused:
- Your sales floor. They’ve been hearing the actual words for years. Ask what people say on the phone before they know the right terminology — that’s your keyword list, already written.
- The search terms report. Not what you bought. What people actually typed to get there.
The common trap
Letting the platform decide who your customer is.
Broad match, automated audience expansion, and hands-off campaign types all make the same offer: you don’t have to know who you’re looking for, we’ll figure it out. It’s an appealing offer when the keyword research is hard and the buying committee is complicated.
But the platform has never met your buyer. It has never sat in a plant, never heard the objection your sales team hears every week, never watched a $200,000 decision stall because one engineer wasn’t convinced. It’s pattern-matching against everyone else’s data, and what it optimizes toward is whatever you told it success looks like. Hand it a vague target and it will confidently find you a large number of the wrong people.
This is the one step that cannot be outsourced — not to an agency, not to a platform, not to a model. Understanding your buyer is the input. Everything downstream is execution, and execution against a wrong input is just expensive.
The quieter version of the same trap: treating search volume as demand. A high-volume generic term is usually the wrong people in bulk. A low-volume application term might be three companies, one of which is your entire year.
The deeper judgment
Intent isn’t a property of the phrase. It’s a property of the person typing it.
Two people search the identical words a week apart. One is comparison-shopping with a signed budget. The other is three months from a conversation with their boss. Same query, same ad, same page — completely different moment. There is no keyword that reliably separates them, which is why “we only bid on high-intent keywords” sounds rigorous and quietly isn’t.
What you can do is stop pretending the phrase told you everything, and build for the range of people behind it. That’s why the page needs more than one way forward, and it’s why the thing you count matters so much: you cannot see intent, so you infer it from what people do next. If what you count is wrong, your picture of who’s searching is wrong too — and you’ll go buy more of them.
So this cell doesn’t resolve on its own. The search terms report is the cheapest customer research you will ever own, and most businesses read it as a billing statement. It’s the one place the market tells you, in its own words, what it thought you were selling.
And underneath it: do not remove expertise from the moments where trust is being formed. Knowing who you’re talking to is not a targeting setting. It’s the expertise.
Where this thread goes
Once you know what they were asking, you owe them an answer that matches.
- A click is a promise — keep it on the page — Earn Trust · Search & Advertising
- What you count trains what you get, including who it finds — Improve With Wisdom, Not Hype · Search & Advertising
- Query → message → page → a useful next step — Connect the Experience · Search & Advertising