Keyword Research: How to Identify the Right Keywords for SEO

A step-by-step framework for identifying keywords that drive qualified organic traffic. The keyword types that matter, the tools we use, the validation process, and the spreadsheet structure that makes the output actionable.

Keyword research is the foundation of every campaign we run. Get it wrong and you’ll spend months optimizing for terms that never convert. Get it right and the rest of the SEO process follows naturally.

This guide covers the framework we use across client and partner agency engagements: how to identify potential keywords, how to validate them, how to assess competition realistically, and how to compile the output into a deliverable that drives the rest of the work.

This is the foundation guide. The companion is our keyword mapping guide, which covers what happens after the research is done.

Why keyword research matters

Keyword research determines what your SEO efforts focus on. It’s the input that drives content strategy, on-page optimization, internal linking, and link building. Bad keyword research wastes everything downstream.

The most expensive mistake we see: months of effort optimizing for keywords that don’t actually map to revenue. The traffic comes, but the conversion doesn’t. A solid research process up front avoids the expensive course-correction six months in.

Keyword types

Four categories that frame how to think about the universe of possible keywords.

Head keywords

Short, general, high-volume terms.

• “Search engine optimization”
• “Plumber”
• “Insurance”
• “Tree removal”

High volume, high competition, fuzzy intent. A search for “plumber” could be someone looking to hire one, someone researching the trade as a career, or someone curious about plumbing-related jokes. Without modifiers, head keywords carry too much intent ambiguity to be reliable conversion drivers.

Head keywords are usually competitive enough that ranking for them takes years of authority building, and the traffic they produce often doesn’t convert as well as more specific queries. Useful as part of a broader strategy, rarely the focus of one.

Long-tail keywords

Extended phrases with focused intent.

• “Link building strategies for B2B companies”
• “Emergency plumber in Halifax”
• “Best whole life insurance for self-employed”
• “Tree removal cost factors”

Lower volume per keyword, but cumulatively long-tail volume often exceeds head-term volume. The intent is sharper, conversion rates are higher, and the competition is meaningfully lower. Most realistic SEO campaigns are won at the long-tail level, with head terms as bonus visibility from broader topical authority.

Informational keywords

Knowledge-seeking queries. The user wants to learn, not buy.

• “How to clean a French press”
• “What is keyword research”
• “Why do leaves change color”

High volume, broad reach, low immediate conversion. Informational content captures top-of-funnel relationships that pay off later when the searcher is ready to buy.

Buying keywords

Action-oriented queries indicating purchase intent.

• “Best French press for cold brew”
• “Buy French press online”
• “[Service] near me”

Lower volume, much higher conversion intent. Maps to product, category, and service pages, not blog content.

For the full framework on intent reading, see our searcher intent guide.

How to find potential keywords

Multiple sources, used in combination. No single source surfaces a complete keyword list.

Source 1: Existing site data

Before any external research, mine what your own site already tells you.

• Google Search Console shows queries already driving impressions and clicks to your site. Sort by impressions to find where you have visibility you can capitalize on. Sort by position to find the page-2 keywords that could move to page 1 with focused work. Filter for high-impression / mid-position keywords for the highest-leverage opportunities.
• Google Analytics shows which pages drive organic conversions. Reverse-engineer which keywords those pages target.
• Internal site search. What are visitors typing into your site search? Often surfaces keywords you’re not yet targeting in your content.

Source 2: Competitor analysis

Pull competitor keyword data through Ahrefs, SEMrush, or Moz. The patterns reveal:

• Keywords competitors rank for that you don’t. Direct opportunities.
• Keywords multiple competitors share. Validated as findable and worth ranking for.
• Their highest-traffic content topics. Topics that generate real organic value in your category.
• Gaps in their content. Topics nobody covers well, where strong content can stand out.

For the deeper competitor analysis methodology, see our keyword mapping guide.

Source 3: Industry knowledge

The seller, the founder, the team that talks to customers daily. Customer service tickets, sales call recordings, FAQ logs. The questions people actually ask. The phrases they use. The terminology that’s industry-standard versus the terminology customers use.

This source consistently surfaces keywords the tool-based research misses because they’re conversational rather than search-volume-significant in isolation.

Source 4: Tool-based discovery

Keyword research tools that suggest related and related-related queries.

• Google Keyword Planner. Free, but data quality is mixed since the volume reporting changed in 2016. Useful primarily for paid search but applicable to organic.
• Ahrefs Keywords Explorer. Strongest combination of suggestion quality, related keyword breadth, and difficulty scoring.
• SEMrush Keyword Magic Tool. Similar to Ahrefs with somewhat different data sources.
• Moz Keyword Explorer. Good for difficulty assessment alongside the others.

Source 5: SERP-driven discovery

Run candidate queries directly in Google. Capture:

• People Also Ask boxes. Related questions Google associates with the query.
• Related searches at the bottom of the SERP.
• Autocomplete suggestions. What Google suggests as you type.
• AI Overview source list. When an AI Overview appears, the cited sources are the kinds of pages currently winning for that query.

Categorizing keywords by SERP results

Once you have a candidate list, segment by what the SERP looks like for each query. This determines what kind of page can actually rank.

Local pack queries. Returns a 3-pack map result (e.g., “lawyers in Chicago”). Local SEO and Google Business Profile dominate.

Direct answer queries. Returns an answer box or AI Overview prominently (e.g., “what is tenant insurance”). Featured snippet optimization or AI Overview citation eligibility matters.

Video-dominant queries. Returns a video carousel (e.g., “how to install a washer”). Video content (YouTube primarily) often ranks better than text.

Shopping-dominant queries. Returns shopping results (e.g., “buy running shoes”). Commercial intent, product pages required.

Image-dominant queries. Returns an image pack early (e.g., “modern kitchen design ideas”). Image SEO matters more than text SEO.

Mixed/standard organic queries. Traditional 10 blue links dominate. Standard on-page and content optimization works.

The page type you’re building has to match what Google ranks. A blog post on a query Google has decided is informational works. The same blog post on a query Google has decided is shopping doesn’t. SERP analysis surfaces this before you commit content effort.

Calculating organic competition

Three methods to assess whether you can realistically rank for a keyword.

Manual SERP inspection

Open the query, study the first 10 results. What do they have in common? Major brand names? Editorial sites you can’t out-publish? Product pages from established retailers? Or a mix where smaller sites are ranking?

If the SERP is dominated by sites with massive authority (Wikipedia, major news outlets, large brands), the keyword is hard to compete on regardless of what tools say. If smaller, focused sites are ranking, the keyword is achievable.

Tool-based difficulty scores

Ahrefs, SEMrush, and Moz all generate keyword difficulty scores. Each uses a slightly different methodology, but all return a 0-to-100 score where higher means harder.

Treat these scores as relative, not absolute. A “30” in your category against your domain’s authority might be hard. The same “30” in another category against a different domain might be easy. The score is most useful for comparing keywords against each other within the same context, not as a universal threshold.

Backlink and authority analysis

Pull the top 10 ranking pages for the keyword. Examine their referring domain count, page authority, and content depth. The pattern shows what it takes to compete.

If the top 10 all have 200+ referring domains and your site has 30, the keyword isn’t winnable in the near term. If the top 10 have 10 to 50 referring domains and your site is in that range, the keyword is achievable.

Calculating traffic and ROI

Once you’ve validated competition, estimate realistic traffic and revenue potential per keyword.

CTR estimates by position (rough benchmarks; actual rates vary by SERP composition):

• Position 1: ~25%
• Position 2: ~15%
• Position 3: ~9%
• Position 4: ~6%
• Position 5: ~4%
• Positions 6-10: 2 to 4% combined
• Page 2: <1%

Note that the introduction of AI Overviews, featured snippets, and local pack results has compressed organic CTR for many queries. The benchmarks above are starting points, not guarantees.

Example calculation:

• Keyword volume: 1,000 monthly searches
• Target position: 2 (~15% CTR)
• Expected visits: 150 monthly
• Site conversion rate: 2%
• Average customer value: $1,000
• Estimated monthly revenue: 150 × 0.02 × $1,000 = $3,000

This calculation has substantial uncertainty (volume estimates are imprecise, CTR varies by SERP, conversion rate varies by intent match), but it forces a reality check on which keywords are worth pursuing. A keyword with 50 monthly searches at a 0.5% conversion rate generating $50 of expected monthly revenue isn’t worth a focused content effort.

The deliverable

The output of keyword research is a spreadsheet, and the spreadsheet structure determines whether the research drives the rest of the work or sits unused.

Minimum required columns:

Column	Purpose
Keyword	The query being researched
Search volume	Monthly volume from your tool of choice
SERP features	Local pack, featured snippet, AI Overview, video, shopping, etc.
Difficulty	Tool-based difficulty score
Intent	Navigational, informational, commercial investigation, transactional
Client relevance	1 to 3 scale: not relevant, somewhat, highly
Notes	Anything else worth capturing per keyword

Optional columns we add depending on project scope:

• Current ranking
• Target page (the keyword mapping output)
• Estimated traffic at target position
• Estimated revenue
• Priority tier (1 to 3)

Client relevance scoring

Critical and underused. The client (or business owner) is in the best position to judge whether a keyword aligns with what they actually sell, who they actually serve, and where their margin actually comes from.

A keyword can have great volume, low difficulty, and clear intent and still be wrong for the business. “Cheap [service]” might rank well, but if the business positions on quality and margin, ranking for “cheap” attracts the wrong customers.

Get client review on every keyword on the list before committing to optimization work. Their input on relevance will shape the priority order in ways the tools can’t.

Common keyword research mistakes

A few patterns we see repeatedly.

• Volume worship. Chasing high-volume head terms that won’t convert and won’t rank within reasonable timelines.
• No competition reality check. Pursuing keywords without examining whether the existing top 10 is actually beatable.
• Intent mismatch. Targeting transactional keywords with blog content or vice versa.
• No client validation. Producing a keyword list in a vacuum and discovering at handoff that half the keywords don’t fit the business.
• Ignoring SERP features. Not noticing that AI Overviews, local pack, or shopping results dominate the SERP and compress organic CTR.
• Set-and-forget. Building the research once and never updating. Keyword landscapes shift; the research should be refreshed at least annually.

How we approach keyword research at SEO Brothers

Keyword research is the first deliverable in every campaign we run. We use Ahrefs as the primary tool, supplemented by Search Console mining and competitor analysis. The output gets reviewed with the client for relevance scoring before mapping work begins.

For partner agencies, we deliver the keyword research spreadsheet as the foundation document for the campaign. The agency reviews with the client, and we move into mapping once the priorities are locked.

If you’ve got a campaign starting and want a tailored keyword research approach, book a call and we’ll walk through how we’d scope it for your situation.