Straight answer: There are eight types of keyword research worth knowing, and most people running a business or a blog only ever practise two: seed keyword research and competitor research. That gap is exactly why so much content gets written and never ranks. Learn all eight, use five of them regularly, and you stop guessing what to write next.
Why "just find keywords with volume" stopped working
I've been doing this since keyword research meant typing a word into the old Google Keyword Tool and squinting at a bar chart. Back then, one type of research was enough because search itself was simple: type a phrase, get ten blue links. Now search includes AI Overviews, Perplexity answers, Reddit threads ranking above brand websites, and people asking ChatGPT instead of Google in the first place. One flavour of keyword research can't cover that. You need several, used for different jobs.
I worked with a client last year, a small HR software company in Manchester, whose content team had been publishing three posts a week for over a year. Traffic was flat. When I audited their process, every single topic had come from one source: a paid tool's "keyword ideas" list, sorted by volume. They had never once looked at what people asked in support tickets, never checked what competitors ranked for that they didn't, never mapped a single buyer journey. They were doing seed research on repeat, forever. We added just two more types, question-based research and commercial-intent research, and within four months two new pages were driving demo bookings, not just visits.
The eight types, in plain terms
1. Seed keyword research
This is the starting point everyone knows: pick a broad topic (say, "invoicing software"), and pull related terms from a tool. It's necessary but it's the shallowest layer. If this is all you ever do, you'll keep finding the same obvious terms your competitors already rank for. Good tools for this stage are covered in this piece on which keyword search tools are worth using, and honestly, which one you pick matters far less than what you do with the list afterwards.
2. Competitor keyword research
You take three to five direct competitors and pull every keyword their pages rank for that yours don't. I do this with a simple gap report: competitor URL, keyword, their position, my client's position (or blank if we don't rank at all). For that HR software client, this single exercise surfaced 42 keywords with real search intent that they had zero content for, including "employee onboarding checklist template," which now brings in around 900 visits a month on its own.
3. Long-tail keyword research
Three-to-six word phrases with lower individual volume but far less competition and much higher conversion rates once someone lands. "Payroll software" gets fought over by giants. "Payroll software for a 15 person company" barely has competition and tells you exactly who's searching and why. If you're wondering how many of these to target on a single page without diluting it, I wrote a full breakdown here: how many keywords should you target on one page.
4. Question-based (intent) research
This is where you pull directly from "People Also Ask," Reddit, Quora, and your own customer support inbox. Real questions, in real people's words. This type consistently outperforms seed research for one reason: it matches how people phrase things when they're stuck, not how marketers phrase things when they're guessing at a keyword.
5. Commercial and transactional intent research
Keywords that signal someone is close to buying: "best," "vs," "pricing," "alternative to," "for small business." These usually have lower volume than informational keywords but a much shorter path to revenue. Most content calendars are stuffed with informational topics and starving for commercial ones, which is backwards if the goal is actual sales rather than just traffic.
6. Local keyword research
If you serve a physical area, "near me" and city-modified terms ("accountant Bristol," "AI consultant Tel Aviv") are their own research category with their own tools, mostly Google Business Profile insights and local pack tracking rather than standard SEO tools.
7. Seasonal and trending keyword research
Terms that spike at predictable times of year (tax deadlines, Black Friday, back to school) or unpredictably around news events. You research these on a rolling basis using Google Trends and by watching what suddenly gets asked about in your niche, not once and done.
8. Zero-volume and low-volume keyword research
Here's the uncomfortable bit most articles on this topic skip entirely: keyword tools massively undercount real search demand, and treating their volume numbers as gospel is a mistake I still see experienced marketers make. Ahrefs, Semrush, and Google Keyword Planner all show "0" or "10" monthly searches for thousands of highly specific, high-intent phrases, particularly anything longer than five words or anything phrased the way a real person types when they're frustrated at 11pm. I've ranked pages for "zero volume" keywords that quietly bring in leads every single month because the tool simply never sampled enough data for that exact phrase. If you only chase what the volume column tells you to chase, you will miss a genuine chunk of your addressable audience. This is doubly true now that AI chatbots answer long, specific, conversational questions that never showed up as searches in the first place.
A simple weekly process that covers all eight without burning out
You don't need to run all eight types every single week. Here's roughly how I split it across a month for a client with one blog and modest resources:
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
- Week 1: Seed and competitor research, refresh the master keyword list.
- Week 2: Pull ten new question-based topics from support tickets and forums.
- Week 3: Check for one seasonal or trending opportunity relevant to the niche.
- Week 4: Review commercial-intent gaps against competitors and long-tail variations for pages already ranking on page two.
That rotation alone gives you fresh input from every category every month, without anyone spending a whole day buried in spreadsheets. It also feeds directly into a wider content plan; if you haven't mapped out what that plan should contain beyond keywords, this guide on what should be included in a solid content marketing strategy is worth reading alongside this one.
The mistake I see most often now
People treat keyword research as a task they do once, at the start of a content project, then never revisit. Search behaviour shifts constantly. A keyword that had zero commercial intent two years ago can turn transactional overnight because a competitor launched a comparison page and started stealing the conversation. Revisit your keyword categories quarterly at minimum. And remember that keyword research on its own does nothing; it has to connect to publishing cadence and to the rest of your site's authority. If you're producing content but wondering whether you're publishing at a pace that matches your research, there's a practical answer in how many blog posts you should publish a day for growth, and if your backlink profile is thin, even perfectly researched keywords will struggle, which is why I also point clients toward how many backlinks a website should realistically have.
Which types matter if you only have time for three
If you're a solo founder or a small team and eight categories feels like too much, prioritise these three: competitor gap research (fastest way to find proven demand), question-based research (matches how real people search and how AI answers work now), and commercial-intent research (the type most directly tied to revenue). Seed research is fine as a foundation, but it should never be the whole strategy, and if you want the fuller case for why bloggers specifically underrate this whole exercise, I laid it out in why keyword research matters for bloggers trying to grow traffic.
If you want the wider picture, see What are search terms and how do you find the ones your customers use.
If that raises the next question, see How to Do Competitor Keyword Research for Your Niche (A Straight Process, Not a Tool List).
You don't need the biggest tool to check this. See these alternatives to Ahrefs, including the free route.
Frequently asked questions
How many types of keyword research are there in total?
Eight are worth knowing in practice: seed, competitor, long-tail, question-based, commercial/transactional, local, seasonal/trending, and zero or low-volume research. Most people only ever use two, seed and competitor, which is why their content plans stall out over time.
Do I need to do all eight types every time I write content?
No. Rotate through them across a month rather than trying to do all eight at once. Competitor, question-based, and commercial-intent research give the fastest return if you only have time for three.
Is search volume the most important number in keyword research?
Not on its own. Volume tools consistently undercount long, specific, conversational phrases, which now matter more than ever because of AI-driven search. A page targeting a "zero volume" keyword can still bring in steady leads month after month.
What's the difference between long-tail and question-based keyword research?
Long-tail research looks for lower-competition phrases regardless of format, often three to six words. Question-based research specifically pulls the actual questions people ask, pulled from forums, support tickets, and "People Also Ask" boxes, which tend to match search and AI answer intent more precisely.
