Gixo Quill · content guide
Keyword Research
Keyword research is deciding which searches are worth writing for. Finding terms is the easy half; the hard half is reading intent correctly and being honest about which pages you can realistically rank.
Last reviewed: August 2026
The parts that actually decide the outcome
- Intent beats volume. A 200-a-month term where the searcher wants to buy is worth more than a 20,000 term where they want a definition.
- Read the results, not the tool. The pages currently ranking tell you what the search engine thinks the query means. If they are all listicles, your guide will not rank.
- Difficulty scores are relative, not absolute. They compare link profiles. They cannot see whether the ranking pages actually answer the query well.
- One page per intent, not per keyword. Variations of the same question belong on one page; splitting them creates cannibalisation.
- Some keywords should not get a page. If you cannot say something the top results do not, the honest answer is to skip it.
Four kinds of intent, and what each needs
| Intent | Query looks like | Page that wins |
|---|---|---|
| Informational | what is, how does, why | Explanation with a direct answer up front |
| Commercial | best, top, vs, alternatives | Comparison with real criteria |
| Transactional | buy, pricing, free trial | Product or pricing page |
| Navigational | a brand name | That brand's own page, usually |
A workable process
Keyword research for SEO: what changes when the goal is ranking
Keyword research for a business and keyword research for SEO are not the same exercise. The first asks what your buyers want to know. The second adds a constraint: whether a search engine will allow your page to be the answer.
- The results become part of the input. Once ranking is the goal, the ten pages already there define the format, the depth and the angle you have to match. This is the step keyword research most often skips entirely.
- Volume is read per cluster, not per phrase. Twenty variants of one question are one page and one opportunity, not twenty of either.
- Your own site becomes a competitor. On an established site, half the keyword decisions are really decisions about which existing page should own the term.
- Difficulty is judged against your site, not in the abstract. The score describes link profiles in general. The useful question is whether any site of roughly your authority appears on the first page for that query at all.
- Some research correctly ends in not writing. A cluster where every ranking page answers the query well and you have nothing to add is a finding, not a failure.
A worked cluster
One seed expanded, grouped by intent, with the keep-or-skip call shown. Seed: expense report.
- expense report template / form / excel. One intent, one page, and the results fix the format: a downloadable artefact, not an article. Keep.
- what is an expense report / meaning / definition. One definitional page, answering in the first forty words. Keep.
- how to fill out an expense report / how to submit one. The same intent as each other and a different one from the two above. One how-to page. Keep.
- expense report software / best expense report app. Commercial intent; the results are review sites and vendor pages. Keep only if you sell one. Otherwise this is where the traffic stops converting.
- expense report fraud statistics. A real query with a real answer, and you have no data of your own. Skip — without original numbers there is nothing to add to what already ranks.
Five clusters from one seed, four pages, one deliberate refusal. The refusal is the part that makes it research rather than a list.
How Gixo Quill handles this
Quill is built around producing the page once the target is chosen, with formats that match intent — WhatIsArticle and HowToGuide for informational, Comparison and AlternativesList for commercial, BuyingGuide for transactional. The deterministic overlap check exists specifically to catch the cannibalisation case before you publish.
Frequently asked questions
What is keyword research?
Deciding which searches are worth writing for, based on what the searcher wants, what already ranks, and whether you can add anything to it.
How do I find keywords?
Start from the questions your buyers actually ask, then use a tool to expand and group them. Tools are good at expansion and poor at knowing your buyer's language.
Is search volume the most important metric?
No. Intent matters more. A low-volume term where the searcher is ready to buy is usually worth more than a high-volume term where they want a definition.
How reliable are keyword difficulty scores?
They compare link profiles and are useful for rough triage. They cannot judge whether the ranking pages actually answer the query well, which is often where the real opportunity is.
Should every keyword get its own page?
No. Group variations that want the same answer onto one page. Splitting them creates pages that compete with each other, which is the most common self-inflicted ranking problem.
How do you do keyword research for SEO?
Start from the questions your buyers actually ask, expand them with a tool, group the variants that want the same answer into one cluster per page, open the results for each cluster to see which format ranks, and drop the clusters where you cannot add anything to what is already there.
Do you need a keyword tool to do keyword research?
Not to do it, only to do it at scale. Tools are good at expansion and volume estimates and poor at knowing your buyer's language or judging whether a ranking page actually answers its query. The judgement half is manual either way.
Related content guides
Turn the chosen keyword into a page that fits its intent
Quill matches the content format to the intent rather than producing the same article shape for every target.
Inside Quill
Review concrete content and SEO signals before delivery
