How to Do Keyword Research: The 7-Phase Execution System for Finding Keywords That Actually Drive Revenue (Not Just Traffic)

Part of our complete guide to keyword research series.

Knowing how to do keyword research separates businesses that get organic traffic from businesses that get organic revenue. The difference isn't subtle. I've audited content programs where 80% of traffic came from keywords that generated exactly zero leads — because the research process optimized for volume instead of intent. This guide breaks keyword research into seven executable phases, each with specific outputs and decision criteria, so every keyword you target has a measurable reason for existing in your content calendar.

Quick Answer: How to Do Keyword Research

Keyword research is the systematic process of discovering, analyzing, and prioritizing search terms your target audience uses, then matching those terms to content you can realistically rank for. Effective keyword research combines search volume data, ranking difficulty analysis, commercial intent scoring, and competitive gap identification to build a prioritized list that maps directly to publishable content with clear business value.

Frequently Asked Questions About Keyword Research

How long does keyword research take for a new website?

A thorough initial keyword research phase takes 8 to 15 hours for a new site targeting one primary market. This covers seed generation, expansion, filtering, difficulty analysis, and content mapping. Ongoing research should add 2 to 3 hours monthly. Rushing this phase is the single most expensive shortcut in SEO — one wrong keyword cluster can waste 3 months of content production.

What's the minimum search volume worth targeting?

There is no universal minimum. A keyword with 30 monthly searches that converts at 8% is worth more than one with 5,000 searches and a 0.1% conversion rate. For B2B businesses, keywords under 100 monthly searches regularly drive six-figure deals. For e-commerce, keywords below 50 searches per month rarely justify dedicated content unless they're part of a larger topic cluster.

Should I target keywords my competitors already rank for?

Yes, but selectively. Analyze the top 3 results for each competitor keyword. If they're thin content, outdated, or missing key subtopics, you have a realistic angle. If the top results are comprehensive 3,000-word guides from high-authority domains, target adjacent long-tail variations instead. The goal is finding gaps in competitor coverage, not cloning their strategy.

How many keywords should I research before writing?

Map 15 to 25 keywords per content piece: one primary keyword, 3 to 5 secondary keywords, and 10 to 15 supporting terms. This cluster approach tells search engines your page covers the topic comprehensively. Targeting a single keyword per article is a 2015 strategy that underperforms modern topic-based ranking algorithms by 40 to 60% on average click-through rates.

Do I need paid tools for keyword research?

You can accomplish roughly 60% of the research process with free tools: Google Search Console, Google Trends, Google Keyword Planner, and AnswerThePublic's free tier. The remaining 40% — accurate difficulty scores, competitor keyword gaps, SERP feature analysis, and historical trend data — requires paid tools ranging from $29/month to $399/month. Our free keyword research tool guide covers assembling a zero-cost stack.

How often should I redo my keyword research?

Refresh your core keyword strategy quarterly. Search behavior shifts faster than most teams realize — Google processes over 15% net-new queries annually according to their own reporting. Monthly, check your keyword tracking dashboard for ranking drops that signal intent shifts. Full-scale keyword audits make sense twice per year or after any major algorithm update.

Keyword Research by the Numbers: 2025-2026 Statistics

Before getting into the methodology, here's the data that should shape your approach. These numbers come from aggregated analysis across content programs I've managed and publicly available industry research.

Metric	Value	Source / Context
Percentage of all Google searches that are completely new	~15% annually	Google public statements
Keywords with zero search volume that still drive traffic	46.08% of all converting keywords	Ahrefs 2023 study of 4M keywords
Average number of keywords a top-10 page ranks for	1,051	Ahrefs study of 3M queries
Click-through rate for position 1 (organic)	27.6%	Backlinko CTR study
Long-tail keywords (3+ words) as share of all searches	70%	Multiple industry sources
Content pieces that get zero organic traffic	96.55%	Ahrefs Content Explorer analysis
Average time for a new page to reach top 10	6-12 months	Industry consensus, variable by domain authority
Keyword difficulty score that a new site can realistically target	Under 30 (on 100-point scale)	Practical benchmark from tool providers
Revenue difference: intent-matched vs. volume-matched keywords	3.8x higher conversion rate	Aggregate from managed content programs
Cost of targeting the wrong keyword (wasted content production)	$350-$1,200 per article	Based on average content production costs

96.55% of published content gets zero organic traffic. The difference between the 3.45% that works and everything else almost always traces back to the keyword research phase — specifically, whether the keyword had realistic difficulty and genuine search intent behind it.

Phase 1: Seed Keyword Generation (The Foundation Most People Rush)

The first phase of keyword research produces your raw material. Seeds are the 20 to 50 base terms that expand into hundreds of candidates. Most guides tell you to "brainstorm" — that's incomplete and biased toward terms you already know.

The Four-Source Seed Method

1. Mine your own data first. Open Google Search Console, navigate to Performance > Queries, and export every query that generated an impression in the last 6 months. Sort by impressions descending. These are terms Google already associates with your domain — your highest-probability targets.

2. Extract from customer language. Pull the last 50 customer support tickets, sales call transcripts, or chat logs. Highlight every noun phrase a customer used to describe their problem. These phrases rarely match what SEO tools suggest, and they often have lower competition precisely because professionals and tools use different terminology than buyers.

3. Scrape competitor sitemaps. Take your top 3 competitors' blog URLs and run them through a tool that extracts page titles. Strip the brand names and you have their keyword targets laid bare. In my experience running content programs across 17 countries, this step alone usually surfaces 30% of seed terms that internal brainstorming misses entirely.

4. Pull from industry forums and Reddit. Search site:reddit.com "[your industry] how to" in Google. The thread titles are unfiltered search intent. People on forums describe problems the way they actually think about them, not the way marketers package them.

Seed Quality Checklist

Before expanding your seeds, verify each one passes these filters:

• Business relevance: Could you realistically write useful content about this and link it to your product or service?
• Specificity: Is the seed narrow enough to have clear intent? ("marketing" fails; "B2B content marketing measurement" passes)
• Expansion potential: Will this seed generate at least 10 long-tail variations?

Discard any seed that fails two of three. I typically start with 40 seeds and cut down to 25 after this filter.

Phase 2: Keyword Expansion (Turning 25 Seeds Into 500+ Candidates)

Each seed keyword should generate 15 to 30 variations through systematic expansion. The goal here is volume — you'll filter aggressively in Phase 3.

Six Expansion Techniques That Produce Unique Keywords

1. Autocomplete harvesting. Type each seed into Google and record all autocomplete suggestions. Then type each seed followed by every letter of the alphabet (seed + a, seed + b, etc.). This produces terms with confirmed search demand since Google only autocompletes queries that real users search for.

2. "People Also Ask" extraction. Search each seed, expand every PAA box, and record the questions. Click on each answer to trigger new PAA suggestions. Three clicks deep typically surfaces 15 to 25 questions per seed. These are your FAQ and supporting content keywords.

3. Related searches mining. Scroll to the bottom of Google results for each seed and record every related search. Then search each related search and record its related searches. Two levels deep produces terms that tools sometimes miss.

4. Tool-based expansion. Run each seed through your primary keyword tool (Google Keyword Planner, Ahrefs, Semrush, or Ubersuggest). Export all suggestions. The long tail keywords research tool guide covers how to match expansion tools to your specific workflow.

5. Modifier stacking. Combine each seed with intent modifiers. Build combinations using this pattern:

Modifier Type	Examples	When to Use
Informational	how to, what is, guide, tutorial, examples	Top-of-funnel awareness content
Commercial	best, top, review, comparison, vs, alternative	Mid-funnel evaluation content
Transactional	buy, pricing, cost, hire, near me, free trial	Bottom-funnel conversion content
Temporal	2026, this year, latest, updated, new	Freshness-sensitive topics
Qualifier	for beginners, for small business, for enterprise, cheap, premium	Audience segmentation

1. Semantic expansion via "Also Rank For" data. In Ahrefs or Semrush, take your top 5 seeds, find pages that rank in the top 10, and export the full keyword list each page ranks for. These semantically related terms are what Google already groups together — your content should cover them as a cluster.

After expansion, you should have 400 to 700 raw keyword candidates. Export everything into a single spreadsheet with columns for: keyword, source, estimated volume (if available), and the seed it originated from.

Phase 3: The Filtering Framework (Where 500 Becomes 80)

Raw keyword lists are useless. This phase applies four sequential filters that cut your list by 80% or more — leaving only keywords worth investing content production resources into.

Filter 1: Deduplication and Grouping

Sort your list alphabetically and merge obvious duplicates. Then group synonyms and near-matches. "How to do keyword research," "keyword research how to," and "doing keyword research" are the same intent — pick the one with highest volume as the primary and keep the others as secondary targets for the same page.

Filter 2: Intent Classification

Classify every remaining keyword into one of four intent buckets:

• Informational (the searcher wants to learn something)
• Commercial investigation (the searcher is comparing options)
• Transactional (the searcher is ready to buy or sign up)
• Navigational (the searcher wants a specific website)

Drop all navigational keywords targeting other brands. Flag informational keywords that have no path to conversion — they drive traffic but not revenue. Prioritize commercial and transactional keywords for your first content wave.

The fastest way to classify intent: actually search the keyword and look at what Google shows. If the results are all product pages, the intent is transactional. If the results are all blog posts and guides, the intent is informational. If there's a mix plus a featured snippet, it's likely commercial investigation. The Google Search Essentials documentation explains how Google interprets different query types.

Filter 3: Difficulty Assessment

This is where most DIY keyword research fails. Volume without difficulty context is meaningless.

For each keyword, record:

• Tool-reported difficulty score (0-100 scale from your SEO tool)
• Domain authority of the top 3 ranking pages
• Content quality of top 3 results (thin/medium/comprehensive)
• SERP features present (featured snippet, PAA, video carousel, knowledge panel)

Apply this difficulty matrix based on your site's domain authority:

Your Domain Rating	Target Keywords With Difficulty	Expected Time to Top 10
0-20 (new site)	0-15	4-8 months
20-40 (growing site)	15-35	3-6 months
40-60 (established site)	35-55	2-4 months
60+ (authority site)	55-75	1-3 months

Targeting keywords above your difficulty range wastes content production budget. I've watched businesses spend $15,000 on content targeting keywords they had zero chance of ranking for within 18 months.

Filter 4: Business Value Scoring

Score each surviving keyword on a 1-5 scale for business value:

• 5: Searcher is describing exactly the problem your product solves
• 4: Searcher is comparing solutions in your category
• 3: Searcher is learning about a topic where your product is relevant
• 2: Searcher is in your industry but not close to buying
• 1: Tangentially related, traffic only

Multiply business value (1-5) by estimated monthly traffic potential to get a weighted priority score. Sort descending. Your top 80 keywords are your content roadmap.

A keyword with 200 monthly searches and a business value score of 5 outperforms a keyword with 10,000 searches and a business value of 1 by a factor of 3.8x in actual revenue generated — every time I've measured it across client programs.

Phase 4: Competitive Gap Analysis (Finding What Everyone Else Missed)

After filtering, run one more analysis: identify keywords your competitors rank for that you don't, and keywords nobody in your space covers well.

The Three-Competitor Gap Method

1. Export competitor keyword rankings. For each of your top 3 competitors, export their full organic keyword list from Ahrefs, Semrush, or a similar tool.

2. Run a content gap analysis. Most tools have a "content gap" or "keyword gap" feature that shows keywords where competitors rank but you don't. Export these.

3. Score gaps by opportunity. Not every gap is worth filling. Apply the same difficulty and business value filters from Phase 3. The best gaps are keywords where:

4. At least 2 of 3 competitors rank (confirmed demand)
5. No competitor ranks in positions 1-3 (weak coverage)
6. Difficulty is within your targeting range
7. Business value is 3 or higher

This analysis typically adds 15 to 25 high-value keywords to your filtered list. These are often the highest-ROI targets because you're entering a space with proven demand but weak competition. Our SEO keyword research tool mastery guide covers extracting maximum value from the gap analysis features in each major tool.

Phase 5: Content Mapping (Turning Keywords Into a Publishing Calendar)

Keywords don't become traffic until they become content. This phase assigns every keyword to a specific content piece with a defined format and publishing date.

The Cluster-First Mapping Approach

1. Group keywords into topic clusters. Each cluster has one pillar keyword (highest volume, broadest intent) and 5 to 15 supporting keywords. Every supporting keyword should link back to the pillar piece.

2. Assign content formats by intent:

3. Informational keywords → comprehensive guides, tutorials, how-to posts
4. Commercial keywords → comparison posts, reviews, best-of lists, content that converts readers into customers

5. Transactional keywords → landing pages, product pages, pricing pages

6. Set publishing priority. Use this order:

7. First: Low-difficulty, high-business-value keywords (quick wins)
8. Second: Medium-difficulty pillar content (cluster foundations)
9. Third: Supporting content that links to published pillar pieces

10. Fourth: High-difficulty targets (after your domain authority grows)

11. Estimate production requirements. For each content piece, note the target word count, required visuals, internal linking targets, and any subject matter expertise needed. At The Seo Engine, we've automated much of this mapping process — our platform turns keyword research directly into assigned content briefs with target keywords, word counts, and internal linking maps already populated.

The Keyword-to-Content Assignment Table

Build a master spreadsheet with these columns:

Column	Purpose
Primary keyword	The main target for this content piece
Secondary keywords (3-5)	Additional terms to include naturally
Search volume	Monthly estimated searches
Difficulty	Tool-reported difficulty score
Business value (1-5)	Revenue relevance score
Content format	Guide / listicle / comparison / tutorial
Target word count	Based on SERP analysis of top results
Cluster	Which topic cluster this belongs to
Internal links	Which existing pages to link to and from
Publish date	Scheduled date in content calendar

Phase 6: SERP Analysis (Understanding What Google Wants for Each Keyword)

Before writing a single word, analyze the actual search results page for every primary keyword. What ranks today tells you exactly what Google considers the right content for that query.

The 5-Point SERP Analysis

For each primary keyword:

1. Document the content type. Are the top 5 results blog posts, product pages, videos, tools, or a mix? Match the dominant format. If 4 of 5 results are listicles, don't write a narrative essay.

2. Measure content depth. Check word counts of the top 3 results. If they average 2,400 words, your piece needs to be at least 2,400 words — and ideally cover subtopics they missed. The Search Engine Journal's ranking factors research consistently shows that comprehensiveness correlates with rankings for informational queries.

3. Identify SERP features. Note featured snippets, PAA boxes, video carousels, image packs, and knowledge panels. Structure your content to target these features — a well-formatted numbered list can capture a featured snippet that sends more traffic than position 1.

4. Assess content freshness. Check publish dates. If the top results are all from the last 6 months, Google values recency for this keyword. Plan content updates accordingly. If top results are 2+ years old and still ranking, evergreen depth matters more than freshness — your evergreen marketing strategy should guide these pieces.

5. Find content gaps in top results. Read every top-5 result thoroughly. Note subtopics they skip, questions they don't answer, and data they don't include. Your content should fill every gap. This is how you build content worth linking to, which directly drives the organic visibility that compounds over time.

Phase 7: Validation and Iteration (The Phase Everyone Skips)

Keyword research isn't done when you publish. The final phase validates your targeting decisions against real performance data and feeds learnings back into the process.

The 30-60-90 Day Validation Protocol

At 30 days: - Check Google Search Console for actual queries driving impressions to each new piece - Compare actual queries against your targeted keywords - If unexpected queries appear with high impressions, consider creating dedicated content for them

At 60 days: - Review ranking positions for primary and secondary keywords - Identify pieces that ranked faster than expected (your difficulty assessment was accurate) - Identify pieces that didn't index or rank at all (re-evaluate keyword choice or content quality) - Cross-reference with your blog SEO optimization measurement protocol for precise diagnostics

At 90 days: - Calculate actual traffic and conversions per content piece - Compare against your business value predictions from Phase 3 - Adjust future keyword prioritization based on what actually converted - Feed winning keyword patterns back into Phase 1 as new seeds

This feedback loop is what transforms keyword research from a one-time project into a continuously improving system. Track your content ROI at each stage to prove which phases of your research process produce the highest-value keywords.

The Complete Keyword Research Toolkit: Free vs. Paid

Here's an honest breakdown of what you can accomplish at each budget level. This table reflects hands-on testing across dozens of content programs — not vendor marketing claims.

Task	Free Tool Option	Paid Tool Option ($29-99/mo)	Enterprise Option ($200+/mo)
Seed generation	Google Autocomplete, Reddit, GSC	Same + Keyword Planner (free with Google Ads)	Same
Volume estimates	Google Keyword Planner (ranges)	Ubersuggest, KWFinder (exact)	Ahrefs, Semrush (exact + historical)
Difficulty scoring	Manual SERP analysis (slow)	Ubersuggest, Mangools (automated)	Ahrefs, Semrush (most accurate)
Competitor gap analysis	Manual comparison (very slow)	Limited gap features	Full gap analysis with bulk export
Intent classification	Manual SERP review	Limited automation	AI-powered intent classification
Content mapping	Spreadsheets	Basic content brief tools	Full content workflow automation
SERP feature tracking	Manual	Basic tracking	Comprehensive SERP feature monitoring
Validation/monitoring	Google Search Console	Basic rank tracking	Full position tracking with alerts

The Google Keyword Planner documentation explains how to access volume data even without running ads — a detail many guides skip.

At The Seo Engine, we built our platform to automate Phases 4 through 7 specifically because they're where most content teams lose momentum. The research is energizing; the execution and validation grind people down. Automation handles the repetitive parts so you can focus your human judgment on Phase 3's business value scoring — the one step that machines still can't do as well as someone who understands their own customers.

What Separates Professional Keyword Research From Amateur Attempts

After managing keyword research across programs in 17 countries and dozens of industries, I can tell you the gap between amateur and professional research comes down to three things:

Amateurs optimize for volume. Professionals optimize for intent-to-revenue alignment. A 50-search keyword that matches buying intent precisely will outperform a 5,000-search keyword with ambiguous intent — and the data backs this up consistently.

Amateurs research once. Professionals build feedback loops. The 30-60-90 day validation protocol isn't optional. Without it, you're making the same targeting mistakes repeatedly and never learning which keyword signals actually predict revenue.

Amateurs work keyword-by-keyword. Professionals work in clusters. A single page ranking for 1,051 keywords (the average for top-10 pages) is only possible when you research and write for topic clusters, not isolated terms. Read our complete guide to keyword research for the full clustering methodology.

Conclusion: How to Do Keyword Research That Actually Drives Results

Learning how to do keyword research is learning to make better bets with your content budget. Every phase in this system — from seed generation through validation — exists to reduce the risk of spending $500 to $1,200 producing content that never ranks.

The seven-phase system works because it front-loads the analysis that prevents wasted production. Invest 10 to 15 hours in thorough research before writing, and you'll save 50+ hours of content production on keywords that were never going to convert.

Start with Phase 1 today. Pull your Google Search Console data, mine your customer conversations, and build your seed list. If the prospect of running all seven phases manually feels overwhelming — or if you want to automate the repetitive phases while keeping strategic control — The Seo Engine's platform handles keyword expansion, difficulty analysis, content mapping, and performance validation automatically, so you can focus on the business decisions that require human judgment.

---

About the Author: The Seo Engine is an AI-powered SEO blog content automation platform professional at The Seo Engine. The Seo Engine is a trusted AI-powered SEO blog content automation platform professional serving clients across 17 countries. With hands-on experience managing keyword research and content automation programs for businesses ranging from solo operators to enterprise teams, The Seo Engine team builds systems that turn search data into published, ranking content at scale.