Keyword Research Checklist: 12 Checks Before You Commit to a Keyword
· Royking Niba
A keyword research checklist is the list of checks a keyword should pass before you build a page around it: the right intent, real demand you have read correctly, a results page you can realistically enter, and no existing page on your own site already doing the job. The twelve checks below are the ones I run, each tied to the data source that answers it. Two things make this version different from the usual list: every tool definition is quoted from the tool’s own documentation, and there is a small dataset pulled from Ahrefs on 8 October 2026 showing how search volume and real traffic potential can point in opposite directions.
The checklist at a glance
| # | Check | Data source | Pass when |
|---|---|---|---|
| 1 | The page’s job is defined | Your own brief | You can say in one sentence what the reader should be able to do after reading |
| 2 | Seeds come from real language | Customers, sales calls, support tickets, forums | At least some seed terms are words your audience used, not words you use |
| 3 | You checked what you already rank for | Search Console Performance report | You know which queries already give you impressions on this topic |
| 4 | The list is expanded | A keyword tool | Variants, questions and long-tail terms are captured |
| 5 | Volume is read as an estimate | Keyword Planner or a third-party tool | You compare volumes as rough sizes, not exact counts |
| 6 | Seasonality is checked | Google Trends | You know whether demand is steady, seasonal or fading |
| 7 | Traffic potential and parent topic are checked | A tool that reports them | The top-ranking page gets meaningful traffic from the whole topic |
| 8 | Intent is confirmed on the live results page | Google itself | The page type you plan matches what already ranks |
| 9 | Difficulty fits your site | Tool difficulty score plus your own judgement | Pages like yours already appear on page one |
| 10 | No cannibalisation | A site search and Search Console | No existing page already targets the same intent |
| 11 | One primary keyword per page | Your keyword map | Secondary terms are mapped, not stuffed |
| 12 | A baseline is recorded | Search Console | You noted impressions and clicks before publishing |
Checks 1 to 4: start from the reader, not the tool
Google’s SEO Starter Guide, last updated 10 December 2025, puts the starting point plainly: “Think about the words that a user might search for to find a piece of your content.” The same guide notes that people who know a topic well search with different words from newcomers, which is why the seed list should come from your audience’s own phrasing before any tool touches it. Then look at what Search Console already shows. Queries where you have impressions but few clicks are often the cheapest wins you will find, because Google has already decided your site is relevant.
Only then expand the list with a tool. The tool’s job is to find variants and questions you missed, not to choose the topic for you.
Checks 5 and 6: read the volume number correctly
Every volume figure is an estimate, and the tools say so. Google’s Keyword Planner help defines average monthly searches as “The average number of times people have searched for a keyword and its close variants”, says that “By default, the number of searches for the term (regardless of language) is averaged over a 12-month period”, and adds that “Your search volume statistics are rounded.” Ahrefs defines its volume as “An estimation of the average monthly number of searches for a keyword over the latest known 12 months of data.” A 12-month average flattens seasonality, so a term with 1,200 searches a month on paper might do 3,000 in its peak month and 300 in its quietest.
That is what Google Trends is for, with one caution. Trends does not show search counts. Its help page says the numbers “are then scaled on a range of 0 to 100 based on a topic’s proportion to all searches on all topics.” A line at 100 means the peak share of searches in the period you chose, not a large absolute number. Use Trends for the shape of demand and a keyword tool for its size.
Check 7: volume and traffic potential can disagree
Ahrefs’ API documentation defines traffic potential as “The sum of organic traffic that the #1 ranking page for your target keyword receives from all the keywords that it ranks for”, and parent topic as the keyword sending the most traffic to the top-ranking page. These two numbers answer a different question from volume: not how often people type this phrase, but how much traffic the winning page actually earns. The table below is real data I pulled from Ahrefs for the United States on 8 October 2026, for keywords in the SEO reporting space.
| Keyword (US) | Monthly volume | Difficulty | Traffic potential | Parent topic | Potential รท volume |
|---|---|---|---|---|---|
| seo metrics | 4,700 | 0 | 1,800 | seo metrics | 0.38 |
| seo reporting | 3,900 | 17 | 450 | reporting seo | 0.12 |
| seo kpis | 1,800 | 5 | 1,700 | kpi seo | 0.94 |
| keyword research checklist | 1,200 | 1 | 450 | keyword research checklist | 0.38 |
| seo keyword ranking report | 1,600 | 24 | 242,000 | google keyword rank checker | 151.25 |
Three readings come out of it. First, “seo reporting” has more than twice the volume of “seo kpis” but roughly a quarter of the traffic potential, so ranked by volume you would pick the weaker target. Second, a ratio far above 1, as with “seo keyword ranking report”, is a warning rather than an opportunity: the parent topic is a rank-checking tool, so the page winning that traffic is a free tool, and an article will not take its place. Third, a ratio well below 1 usually means the top result ranks for few other terms, which often signals a narrow or confused results page. Neither number is the truth. Traffic potential is one page’s estimated traffic from one tool’s index, and it changes when the top result changes. The point of the check is that you look at both before choosing.
Checks 8 and 9: the results page decides
Search the keyword yourself in a clean browser. If the first page is all product pages and you planned an article, or all tools and you planned a guide, the keyword is telling you what Google thinks the searcher wants. That check overrides any number. Difficulty scores help you sort a long list, but read the definition: Ahrefs calls keyword difficulty “An estimation of how hard it is to rank in the top 10 organic search results for a keyword on a 100-point scale.” It is an estimate built mostly from the links pointing at the current top ten. A low score with a results page full of household-name brands is still a hard keyword for a small site.
Checks 10 to 12: protect the pages you already have
Before you write, search your own site for the topic and check Search Console for queries where two of your pages take turns ranking. Two pages chasing one intent split clicks and links, and the fix later is a merge and a redirect. Map one primary keyword to the page, and list the secondary terms the page should naturally cover. Do not repeat them to hit a count. The Starter Guide is direct about this: “Excessively repeating the same words over and over (even in variations) is tiring for users”, and keyword stuffing sits inside Google’s spam policies. The same guide also retires two old habits: “the keywords in the name of the domain (or URL path) alone have hardly any effect” and “Google Search doesn’t use the keywords meta tag.”
Finally, record the baseline. Note the impressions and clicks Search Console shows for the target query group before you publish, then add that group to your fixed reporting set. My guide to SEO metrics explains why a fixed query group is the only honest way to measure ranking movement, and how average position can fall while traffic grows.
Common mistakes this checklist catches
- Choosing by volume alone. The table above shows how that picks a weaker target.
- Writing an article for a tool-intent keyword. If the parent topic is a tool, the results page wants a tool.
- Ignoring your own site. A new page on an existing topic often competes with the old one. My post on internal linking for SEO covers how to point authority at the page you want to rank.
- Treating difficulty as a verdict. It is a sorting aid. The live results page is the verdict.
- Building pages for every variant. Near-identical pages for each phrasing can drift into the territory Google describes as doorway pages, and thin variants are a common cause of index bloat.
If a site has already lost traffic and you are researching keywords to rebuild it, start one step earlier. Find out why it fell first, using my Google penalty recovery guide, because new keywords will not fix a site-wide quality or links problem.
Leave a Reply