Intent decides everything else
A keyword is a phrase someone typed. Intent is what they wanted when they typed it. Getting the phrase right and the intent wrong produces a page that's about the topic and is never shown, because the engine has already decided what kind of result that query deserves.
| Intent | The query looks like | What ranks |
|---|---|---|
| Informational | "what is an api" | Explanations, guides, definitions |
| Navigational | "navital labs login" | The specific site being asked for |
| Commercial | "best itinerary software" | Comparisons, reviews, roundups |
| Transactional | "buy itinerary software" | Product and pricing pages |
This is also the fastest research method available: search the phrase and look at what's already there. The first page is the engine telling you, explicitly, what format it believes that query wants.
Finding phrases without a tool
- Autocomplete — start typing your topic and read the suggestions; those are real, frequent queries
- "People also ask" and related searches — the adjacent questions, handed to you
- Your own search console — the phrases you already appear for, including ones you never targeted, are the highest-signal list you have
- Support tickets and sales calls — the exact words customers use, which are rarely your internal terms
- Community threads — a question asked repeatedly with no clean answer is an unserved query
Paid tools add volume estimates and difficulty scores. They're useful at scale and are not a prerequisite. The estimates are modelled, not measured, and treating them as facts is a common way to talk yourself out of a good page.
Choosing what's worth writing
The instinct is to target the biggest phrase in your field. That phrase is usually contested by sites with a decade of links, and losing to them costs the same effort as winning something smaller.
| "travel software" | "how to write a client itinerary" | |
|---|---|---|
| Volume | High | Low |
| Intent | Ambiguous | Clear and informational |
| Competition | Established, well-linked sites | Mostly thin content |
| Who searches it | Everyone, including students | Someone doing the job you serve |
| Realistic outcome | Nothing, for years | A page that ranks and converts |
Specific phrases have lower volume and dramatically higher relevance. Ten pages each answering one narrow question will out-perform one page attempting the broad term — and they compound, because each is a real answer someone can link to.
One page, one query family
Group phrases that mean the same thing and serve them with one page. "What is an MVP", "MVP meaning", and "minimum viable product definition" are one page. "How to scope an MVP" is a different page, because the intent shifts from understanding to doing.
A research pass in an afternoon
- List the ten questions customers ask before buyingIn their words, from real conversations. This beats any tool as a starting list.
- Search each one and note the dominant formatGuide, comparison, product page, video. That format is now a requirement, not a choice.
- Pull your search console's page-two queriesAnything ranking 11–30 is a page worth improving before writing anything new.
- Group the list into query familiesOne page per family. Where two families feel identical, merge them.
- Write the narrowest ones firstThey're winnable, they're specific, and specific pages attract the links that eventually make broad terms possible.
Common mistakes
- Targeting a phrase whose results are all a format you're not producing
- Chasing the broadest term in the category as a first move
- Two pages competing for the same intent
- Treating tool volume estimates as measurements
- Writing for a phrase nobody in your actual market types, because it looked good in a spreadsheet
Key takeaways
- Intent decides format, and format decides whether ranking is possible at all
- The first page of results is free research on what the engine wants
- Your own search console is the highest-signal keyword source you have
- One page per query family; narrow phrases first because they're winnable
Try it yourself
Take the question your customers ask most often, search it, and write down the format of the top five results. If your planned page isn't that format, change the page — not the phrase.