The Keyword Research Strategy Library
Eight demand-discovery strategies drawn from interviews with working SEOs, each with a workflow and an Role SEO MCP prompt for its data-backed steps.
How to do keyword research through demand discovery
Most guides teach you to export a volume report and sort descending. These strategies start earlier, where demand originates: customer language, question mining, your own Search Console. They end with pages mapped to intent, not keywords stuffed into paragraphs. Each strategy includes a walkthrough and a copy-paste MCP prompt for its data-backed steps.
Seed from conversation, not a volume report
Harvest seed keywords from sales calls and support tickets using the language customers already use.
02What are long-tail keywords, and how to mine them
Find long-tail keywords in People Also Ask, autocomplete, and Search Console queries where your pages already rank.
03Search-intent mapping (hot / warm / cold)
Sort keywords by buying temperature before you write, then build high-intent pages first.
04Cluster keywords into topical hubs
Group keywords by intent and build topical hubs without creating competing pages.
05Programmatic discovery with Search Console
Use MCP to find Search Console queries and pages with room to gain more clicks.
06Opportunity sizing & forecasting
Estimate a cluster's difficulty, traffic range, and payback scenarios before you invest.
07Intent beyond Google (Pinterest, AI, LinkedIn)
Research demand on Pinterest, LinkedIn, and AI assistants.
08Map positioning to real demand
Check whether your category language matches the terms customers search for.
How to supplement Google Keyword Planner
Keyword Planner is built for planning search-ad campaigns. Its historical metrics combine close variants and report approximate monthly searches, so treat them as directional rather than a complete content-discovery dataset.
The strategies in this library supplement it with three sources you can inspect: your customers' language (strategy 01), Google's own question surfaces (strategy 02), and your Search Console reality (strategy 05). Volume data still matters, but it is most useful after you understand the customer language and search intent.
Free sources for keyword discovery
Google's free surfaces (autocomplete, People Also Ask) plus your own Search Console support the discovery. You can run the data-backed parts of these workflows with Role SEO's keyword research and your connected Search Console. Role SEO's MCP lets a compatible AI assistant query both sources while it works through the workflow. Quality SEO data is why the big suites run £80/month and up; Role SEO's paid plan starts at £15/month with monthly credits included, plus top-ups when you need more. You can start for free.
Keyword research FAQ
How do you do keyword research for SEO?
Seed from customer language, expand into long-tails and questions, label by intent, cluster into one-page-per-intent hubs, then validate against Search Console. Volume is the final filter, not the starting point.
How do you do keyword research for free?
The discovery half runs on sources you already have: customer conversations, Google's autocomplete and People Also Ask, and your Search Console. Quality SEO data costs money, which is why the big SEO suites run $100/month and up. You can start Role SEO for free; the paid plan starts at $10/month and includes $10 of usage. If you need more, you can buy top-up credits.
Can you do keyword research without Google Keyword Planner?
Yes. Keyword Planner combines close variants and reports approximate search-volume data designed for ad planning. Use it to sanity-check commercial value rather than as your only source for discovering topics.
What are the 3 types of keywords?
There is no universal set of three. Role SEO uses four intent types: informational, navigational, commercial, and transactional. By shape, terms are often grouped as head, mid-tail, and long-tail.
How do you do keyword research for a blog?
Blogs win in the tail: mine questions, cluster them into topical hubs, and let each post own one question-intent completely rather than skimming ten.
