Keyword research

Keyword research is the practice of finding the exact words and phrases people type into search engines, then understanding why they typed them. The "why" is what separates a ranking page from an invisible one.

Most guides stop at "find keywords with high volume and low competition." That's the 2015 version of keyword research. It produced keyword-stuffed posts that ranked briefly, then got overtaken by content that actually matched what searchers wanted. In 2026 the job is two things: find what people search for and match the intent behind each search.

What keyword research actually does

A keyword list without intent analysis is just a list. Before you write a single word, you need to know:

What format does the searcher expect? (Guide, listicle, tool, comparison, product page)
Where are they in the buying process? (Researching, evaluating, ready to buy)
What's the specific question they want answered?

Google's algorithm has spent the last decade getting better at inferring all three from the search query alone. Your job is to reverse-engineer that inference and build content that fits it perfectly.

The practical payoff: a page that matches intent ranks faster, holds position longer, and converts at a higher rate than a page built around volume alone.

The keyword research workflow

Every solid keyword research process follows five stages, regardless of which tools you use.

Stage 1 — Seed keywords

Start with the obvious terms: what does your site do, what problems does it solve, what would your customer type if they'd never heard of you? Write 10–20 seed terms. Don't filter yet.

Sources for seeds: your own product features, competitor homepage copy, industry jargon, customer support questions, sales call transcripts.

Stage 2 — Expand

Take each seed term and generate variations. The free methods — Google Autocomplete, People Also Ask, related searches at the bottom of a SERP — produce hundreds of variations per seed. If you have GSC data, pull your existing impressions data to find queries you're already appearing for but not converting.

Stage 3 — Cluster

Group expanded keywords by intent, not just by topic similarity. Two keywords can look identical on paper but represent completely different intents:

"email marketing software" → informational, comparing options
"email marketing software for ecommerce" → commercial investigation, closer to buying

Keywords with the same intent belong on the same page. Keywords with different intents need separate pages. Getting this wrong causes keyword cannibalization — two pages competing against each other for the same slot.

Stage 4 — Map

Assign each cluster to a page. Existing pages get a cluster if the content already fits; new clusters become a content backlog. This is keyword mapping, and it's the step most solo site owners skip. Skipping it is why cannibalization happens in the first place.

Stage 5 — Validate

Before you write, open the SERP for your primary keyword and look at the top 10 results. What format are they? What's the average length? Are results dominated by product pages, comparison posts, or how-to guides? That's your brief.

The modern shift: intent mapping over volume chasing

Volume-first keyword research made sense when Google's algorithm was weaker. Pick a term with 10,000 monthly searches, write something halfway decent, and you'd rank. That era ended around 2019 with the BERT rollout.

Today Google understands synonyms, context, and user intent well enough that:

You don't need the exact-match keyword to rank for it.
Content that satisfies the underlying intent outranks content that merely contains the keyword.
Volume estimates from third-party tools are often wrong by 40–60% (Ahrefs and Semrush both pull from clickstream panels, not from Google directly).

The practical implication: stop choosing keywords by volume and start choosing them by intent fit. A 200 searches/month keyword where you can fully satisfy the intent is worth more than a 2,000 searches/month keyword where the SERP is owned by domain-authority giants and your format doesn't match what Google is showing.

Why zero-volume keywords outconvert head terms

Tools like Ahrefs and Semrush round search volume to the nearest data point in their clickstream sample. Anything below a certain threshold rounds to zero — but the traffic is real.

More importantly, zero-volume keywords are almost always longer and more specific. Longer queries mean the searcher has already done their research and is closer to a decision. A page ranking for "email marketing platform for Shopify stores under 1000 subscribers" converts at a higher rate than a page ranking for "email marketing" — even though the second term has 100× the reported volume.

The rule of thumb: 10 zero-volume keywords covering related specifics will collectively drive as much traffic as one mid-volume keyword, and they'll convert better.

How to do keyword research with zero budget

You don't need Ahrefs ($99/mo) or Semrush ($120/mo) to build a solid keyword strategy. The $0 stack:

Google Search Console — Your best source of real, site-specific query data. GSC shows you every query that triggered an impression in the last 16 months, with actual click and CTR data. No sampling, no estimates — this is the real thing. The GSC-only keyword research workflow covers the five-step process in detail.

Google Autocomplete — Every suggestion Google shows as you type is a keyword with real search volume behind it. The alphabet soup method systematically extracts all variations using every letter of the alphabet as a modifier.

People Also Ask — PAA boxes are Google's public display of related question clusters. Each click expands the box and reveals new questions. Mining PAA for 20 minutes produces a content calendar's worth of topics.

Reddit and Quora — Real users describing real problems in their own language. Search site:reddit.com <your topic> and read thread titles for phrasing you'd never generate yourself.

Competitor sitemaps — yourcompetitor.com/sitemap.xml shows every page they've built. Sort by topic and you have their content strategy.

Full breakdown: keyword research without any tools.

Understanding intent types

Every search query fits into one of four intent categories. Getting the category wrong means building the wrong page.

Informational — The searcher wants to learn. "How does compound interest work." "What is keyword cannibalization." Format: guides, definitions, how-tos.

Navigational — The searcher wants a specific site or page. "Gmail login." "RankCrab pricing." Format: don't try to rank for competitor navigational queries — you won't win.

Commercial investigation — The searcher is evaluating options before buying. "Best project management software." "Ahrefs vs Semrush." "Clearscope alternatives." Format: comparison posts, reviews, roundups. See commercial vs transactional intent.

Transactional — The searcher is ready to act. "Buy standing desk." "Ahrefs discount code." "Sign up Mailchimp." Format: product or landing pages with clear CTAs.

Misaligning format to intent is the most common reason a well-researched keyword fails to rank. The intent mismatch guide walks through the diagnostic and fix.

The LSI keyword myth

A note on "LSI keywords" — a term that still circulates in SEO content despite being based on a misunderstanding. LSI (Latent Semantic Indexing) is a document-retrieval technique from the 1980s. Google's John Mueller confirmed in 2019 that Google does not use LSI.

What Google does use: neural embeddings (via BERT and later models) that understand relationships between concepts, not just keyword co-occurrence. The practical difference is that you can't "trick" Google's algorithm by stuffing in a list of LSI synonyms. What works is covering the topic completely — including the questions and subtopics that appear in the top-ranking SERP results.

Full breakdown: LSI keywords vs semantic keywords.

Keyword mapping: the step that holds the strategy together

Once you have a clustered keyword list, every cluster needs a home. Keyword mapping is the process of assigning clusters to pages in a structured document.

A simple mapping spreadsheet has five columns: primary keyword, intent type, supporting keywords, target URL, status (existing/planned/live). This document does three things:

Prevents cannibalization by making explicit which page owns which cluster.
Reveals content gaps — clusters with no assigned page are your next articles to write.
Provides a shared reference if you're working with a writer or agency.

How to map keywords to pages covers the template and assignment rules.

All guides in this cluster

RankCrab and keyword research

RankCrab's audit and content optimizer don't replace keyword research — they pick up where it ends. Once you've identified your target keyword and mapped intent, RankCrab's 80-check audit tells you exactly what's missing from the page you're optimizing, and the rank tracker shows whether your positioning is moving. It's built for bloggers, indie developers, and small agencies who want the signal without the $500/mo overhead. Plans start at $29/mo, and the free tools (keyword density analyzer, meta tag previewer, internal link extractor) are available without an account.

Free tool

Keyword density analyzer

Spot the phrase you're leaning on too hard before Google does.

Try it