Welders search by brand name more than I expected. I had to retrofit brand names and manufacturer equivalents into existing pages after seeing the data.
That's from my GageRef case study, my welding and fittings reference site. A welder who types a brand and a rod number already knows what they need. They're picking which one to get and making sure it's the right one. That's commercial intent, and it has none of the words my own search intent guide tells you to look for.
What does commercial intent look like in a search?
Someone with commercial intent is choosing: which product, which size, which option fits, before they buy. It shows up in two kinds of searches, and they look nothing alike.
| What they're doing | What they type | What the page has to do |
|---|---|---|
| Comparing options | “best,” “top,” “vs,” “review,” “difference between” | Lay the options side by side with honest trade-offs |
| Picking the exact thing | The product, the size, the spec or the brand | Confirm it's the right one and show what to check before buying |
The first kind is easy to filter for. The second is harder to spot, and on my sites it's where most of the buying searches turned out to be. What a good comparison page needs is in my post on commercial search intent; this one is about finding the people who need it.
How do you filter Search Console for commercial intent?
Start with the four words. In Search Console, open Performance, then Search results. Add a filter, choose Query, then Custom (regex), and paste this:
\b(best|top|vs|review)
That keeps every search with “best,” “top,” “vs” or “review” in it. With the filter still on, click the Pages tab above the table to see which of your pages those searches land on. Those are the people comparing, and it's a good list to have.
It's also the easy half.
Why does a “best” and “vs” filter miss most buyers?
Because by the time people have picked, they stop comparing and start naming the thing.
I could test that on my own sites. Bing Webmaster Tools exports the searches its AI runs before it cites a page, and since June it labels each one with an intent. So I had Bing's own label for every search and could hold my four words up against it. You don't need Bing's data to use what follows; it's how I found the gap. On Comparison searches, the four words did fine: they found 80 of 112. On Commercial searches they found 20 of 113. 🫠
The ones they missed look ordinary until you picture who's typing:
- “40 amp circuit breaker”, on WireRef, my electrical reference site. They know the amperage. They're picking the breaker.
- “hardface welding rod”, on GageRef. They know the job needs a hardfacing rod. What's left is which one to buy.
- “water softener salt crystals”, on CheckMyTap, my tap water site. No “best” anywhere, and they're about to buy a bag of salt.
On my sites, the closer someone got to buying, the more specific the search became and the fewer comparison words it used. The full count is in the table at the end.
How do you find the searches that name the thing?
Build a second filter from your own nouns: the products, sizes, specs and brands your pages help people pick. Nobody else's list works here, because the words come from what you sell or explain. For WireRef it's breakers, wire, cable and amps. For GageRef it's rod numbers and brand names, which is exactly what I had to add after the fact.
Type yours below and copy the filters into Search Console. It runs on this page; nothing you type is sent anywhere.
1. People comparing
\b(best|top|vs|review)2. People naming the thing
\b(breaker|wire|cable|amp)3. Comparing in a full sentence
\b(difference|between|which|better)\bWhat about comparisons that never say “vs”?
People compare in full sentences, too. Bing labeled “difference between single pole and double pole circuit breaker” a Comparison, and none of the four words are in it. The third filter above catches that shape: “difference,” “between,” “which” and “better.” Expect some noise from it, because “which” also starts plenty of learning questions, so read the list before you act on it.
How do you know the filter found real buyers?
Search a few of them yourself and look at what Google shows. The results page is Google's best guess at what most people typing that search want.
| What the results page shows | What it tells you |
|---|---|
| Products, prices, stores and comparison pages | Google reads it as a buying search |
| Definitions, how-to guides and forum threads | Most people typing it want to learn first |
| Some of each | Two kinds of reader type it, so check which pages earn the clicks |
Bing's label is a guess too, and it can be wrong about one person. Someone searching a breaker size might be checking a spec before a job, with no plan to buy today. That's fine. The page that serves them both gives the answer first and then what to check before buying. My post on turning a question into a decision shows how.
Questions people ask about finding commercial intent
Is there a free commercial intent tool?
Search Console is the free one, and it's the best source, because the searches are your own. The two filters above do the sorting, and the builder writes the second filter from your own words. Keyword tools label intent too, but they label searches from their own databases, and the product and spec searches your buyers type are often too specific to show up there.
Can I ask AI to sort my searches by intent?
Yes, as a first pass. Export the queries from Search Console, ask for a label on each with a one-line reason, then check a sample against the live results page. A label is a guess, whether it comes from a model or from Bing, and the results page shows what Google thinks the searcher wants.
What's the difference between commercial and transactional intent?
Commercial means the person is still choosing. Transactional means they've chosen and want to act: buy, book, get a price. The page for the first helps them pick; the page for the second gets out of the way. All four types, with the page each one needs, are in my search intent guide.
What should you do with the list?
Three moves, in this order:
- Find the buying searches that land on a page built to teach. That's the gap that costs you: someone ready to pick gets a lesson. The opposite mistake, a page that sells to someone who isn't ready, is in why transactional pages fail.
- Add what they need to choose to the page they already land on. Sizes, specs, brand equivalents and what to check. That's the retrofit I did on GageRef, and my page brief turns it into a list a writer can build from.
- Build a new page only when no existing page can hold it. The free lesson Is this really a content gap? practices exactly that call on a real report.
The test I use on every page those searches land on: “This page helps someone who searched [X] decide [Y].” If you can’t finish that sentence, the page needs work.
First published November 20, 2026.