Keywords Are Not Pokémon: You Do Not Need to Catch Them All

A founder opens a keyword tool, sees 18,742 possible search terms, and immediately develops the expression of someone who has accidentally become responsible for a small country.

“Should we rank for all of these?”

No. You should also not eat every item in the supermarket because you own a refrigerator.

Keyword research is not a scavenger hunt. It is a decision-making exercise. The goal is not to collect the largest possible spreadsheet, color-code it until it resembles a stained-glass window, and quietly ignore it forever. The goal is to identify the searches that can bring the right people to the right page at the right stage of their buying journey.

That means accepting a liberating truth: most keywords are not your keywords.

The keyword-hoarding problem nobody puts in the pitch deck

Keyword tools are excellent at producing numbers and terrible at producing restraint.

Type in something broad like “project management software” and you will receive:

• Thousands of variations
• Questions from people with wildly different needs
• Competitor names
• Searches from students, job seekers, researchers, and mildly confused humans
• Terms with impressive volume but no obvious commercial relevance
• A handful of gems buried beneath the digital equivalent of attic clutter

The common founder mistake is treating search volume as a scoreboard. A keyword with 20,000 monthly searches appears more exciting than one with 200 searches. But traffic is not revenue. Sometimes traffic is not even qualified traffic. It is merely a crowd standing outside your shop asking where the nearest bathroom is.

Imagine you sell payroll software for restaurants.

Which term is more valuable?

1. “payroll software”
2. “restaurant payroll software with tip pooling”

The first may have more volume. The second tells you the searcher has a specific problem, a relevant industry, and enough operational pain to type several extra words into Google. That person is not browsing. That person is one double-shift away from launching a spreadsheet into the sun.

A keyword is only useful when it matches three things

Before pursuing a keyword, test it against:

1. Audience fit: Is the searcher plausibly your customer?
2. Intent fit: Does the query suggest they need the content or product you offer?
3. Business fit: Can you serve that need well, and does it connect to your growth model?

If any of those are weak, the keyword may still be interesting. It just should not be allowed to commandeer your content calendar like an intern who discovered pivot tables.

Search volume is a clue, not a commandment

Search volume estimates are helpful, but they are estimates. Tools model demand from imperfect data, often group close variants together, and may not capture shifts in behavior, seasonality, geography, or emerging language.

So treat volume as directional evidence, not divine instruction etched onto a tablet.

A keyword with “zero” reported volume may still be valuable if:

• It describes a high-value use case
• It is newly emerging
• It is a specific pain point your sales team hears constantly
• It has buying intent
• It fits a growing product category
• It appears in Search Console impressions already

For a B2B company, ten visits from people evaluating a serious purchase can beat 1,000 visits from people looking for a definition they could get from a dictionary and a cup of tea.

The revenue-weighted keyword question

Instead of asking, “How much traffic can this term bring?”

Ask:

“If we ranked well for this, what would those visitors be trying to accomplish, and would we be glad they landed on our site?”

That question is a remarkably effective nonsense detector.

For example, a company selling customer support software may notice:

The last two can generate traffic. They can also generate a beautifully irrelevant analytics chart.

Intent is the bouncer at the keyword club

A keyword can look relevant while hiding an entirely different intention.

Take “invoice template.” A company selling accounting software might see enormous potential. But search results could be dominated by downloadable templates, free PDFs, and people who want to avoid buying software with the determination of a Victorian miser.

If your page is a sales pitch for invoicing software, Google may reasonably decide it does not satisfy that query. The searcher asked for a template; you handed them a product tour and a form asking for their work email. This is not a match. This is a hostage negotiation.

Read the search results before writing anything

The search engine results page (SERP) is not merely a leaderboard. It is a behavioral report.

For important keywords, search them yourself and look for patterns:

• Are the top results product pages, blog posts, comparison pages, tools, videos, or forums?
• Do results feature “People also ask” questions?
• Are large brands dominating, or are niche sites winning?
• Does Google show local listings, shopping results, image packs, or AI-generated summaries?
• Is the query informational, commercial, navigational, or transactional?
• What does the wording of ranking pages reveal about the searcher’s real problem?

Suppose you sell inventory software and want to rank for “inventory management.”

If the results are mostly broad explainers, guides, and definitions, you likely need a foundational educational page. If results are primarily software category pages with pricing grids and comparison tables, a product-led landing page may be appropriate. If the page is full of “best inventory management software” listicles, you may need a comparison asset or evidence that earns inclusion in those lists.

The SERP tells you the dress code. Do not show up in a tuxedo when everyone else is wearing hiking boots.

Match page type to search intent

A simple working map:

Trying to rank a single “ultimate” blog post for every intent is how websites become content warehouses with no plumbing.

Build keyword clusters, not a keyword graveyard

One page does not need one keyword. Nor does every keyword need its own page.

The useful unit is usually a topic cluster: a primary query, close variations, related subtopics, and supporting pages that answer adjacent needs.

Consider a startup offering employee onboarding software. A sensible cluster might look like this:

Core commercial page

Primary topic: Employee onboarding software

Include natural variations such as:

• onboarding platform
• employee onboarding tool
• new hire onboarding software
• HR onboarding software

These phrases may differ slightly, but they often share a core intent: evaluate software for onboarding employees.

Supporting educational content

Build pages around distinct questions, such as:

• Employee onboarding checklist
• 30-60-90 day onboarding plan
• Remote employee onboarding ideas
• New hire paperwork checklist
• How to measure onboarding success

These pages attract earlier-stage searchers and create useful internal linking paths toward your product pages.

Avoid accidental keyword cannibalization

Cannibalization is not a mystical SEO curse. It usually means two or more pages compete for the same search intent because nobody made a clear editorial decision.

For example, these two articles may be too similar:

• “How to Improve Employee Onboarding”
• “Best Practices for Better Employee Onboarding”

They could be consolidated into one stronger guide unless they serve genuinely different audiences or intents.

Before publishing a new page, ask:

• Do we already have a page that answers this query?
• Is the new page meaningfully different?
• Should this be a section added to an existing page instead?
• Which URL do we want Google and users to see for this topic?

Your website should resemble a well-organized store, not a garage where every box is labeled “important stuff.”

Prioritize by opportunity, not by ego

Founders often gravitate toward glamorous head terms: “CRM,” “marketing automation,” “cybersecurity,” “accounting software.” These keywords are attractive because they are large, obvious, and featured prominently in every competitor’s board deck.

They are also frequently expensive, competitive, and too broad for an early-stage company to win quickly.

That does not mean you should never target them. It means they belong in a long-term strategy, not necessarily next Tuesday’s content brief.

Use a practical opportunity score

You do not need a 47-column model built by someone who says “leveraging” in normal conversation. Score candidate topics from 1 to 5 across a few useful dimensions:

• Customer value: How likely is this audience to become a qualified lead or customer?
• Intent strength: How close is the searcher to a relevant decision?
• Product fit: Can your product genuinely help?
• Ranking feasibility: Can you create something more useful or more credible than what ranks now?
• Content leverage: Can this page support other pages, sales conversations, or email campaigns?
• Evidence advantage: Do you have original experience, data, customer stories, or subject-matter expertise?

A topic with moderate volume, strong intent, and a credible point of view often beats a giant keyword where your best contribution is “we also have a blog post.”

The founder’s unfair advantage: actual proximity to the problem

Big publishers can produce polished summaries. What they often cannot produce is firsthand operational knowledge.

If you have:

• Customer onboarding data
• Support ticket patterns
• Internal benchmarks
• Product usage insights
• Practitioner interviews
• A repeatable implementation process
• Stories from a specific industry

…you have material that can make content substantially better than generic articles.

A cybersecurity startup serving hospitals, for example, should not try to out-Wikipedia Wikipedia on “what is cybersecurity.” It can win trust with practical content such as:

• A ransomware response checklist for small clinics
• HIPAA-friendly incident communication templates
• Questions hospital IT teams should ask before choosing endpoint protection
• A guide to reducing alert fatigue in healthcare security operations

Specificity is not a limitation. It is often the whole meal.

Let customer language do the heavy lifting

The finest keyword research tool in the world cannot replace listening to customers describe their problems in their own slightly chaotic language.

Your audience will tell you what they search for—though rarely in a neat exportable CSV.

Mine these sources:

• Sales call transcripts
• Demo requests
• Support tickets
• Customer interviews
• Product reviews
• Community discussions
• Reddit threads and niche forums
• Internal site search
• Google Search Console queries
• Competitor reviews and comparison conversations

Pay special attention to phrases that reveal pain, urgency, and context:

• “We’re currently using spreadsheets for…”
• “How do we handle this when…”
• “We need a way to…”
• “Our team keeps struggling with…”
• “Is there software that can…”

A founder selling compliance software might discover that prospects rarely search for the category term. They may search for:

• “vendor security questionnaire automation”
• “SOC 2 evidence collection”
• “how to track security policy acknowledgements”
• “GRC software for small SaaS companies”

These are not merely keywords. They are product positioning clues wearing trench coats.

Search Console is your least dramatic source of truth

Once your site has organic visibility, Search Console can reveal:

• Queries where you already get impressions but rank poorly
• Pages attracting unexpected search demand
• Terms where improving a page could move you from page two to page one
• Topic gaps between what you publish and what users seek

Look for pages ranking roughly in positions 5 to 20 for valuable, relevant queries. These are often better optimization opportunities than launching a brand-new article into the wilderness with a packed lunch and no map.

Improve the page by strengthening intent match, clarifying structure, adding missing subtopics, improving examples, and making the page easier to use. Do not simply sprinkle the keyword throughout the copy like parmesan over a bad pasta dish.

Stop chasing keywords that cannot love you back

Some keywords are strategically wrong even if you could rank for them.

Walk away when:

• The searcher wants something you do not offer
• The SERP is dominated by a format you cannot reasonably match
• The query is too broad to produce qualified demand
• The term is mostly navigational for another brand
• Your content would be a weak imitation of existing results
• The traffic has no credible connection to your business
• The topic would distract your team from higher-value work

This is not defeat. It is focus.

A tiny team has finite writing, design, engineering, and subject-matter resources. Every keyword you pursue has an opportunity cost. The best SEO strategy is not the one with the most pages. It is the one where each page has a job.

That job might be to:

• Capture high-intent demand
• Educate an ideal buyer
• Support a sales objection
• Build topical authority
• Earn links because it is genuinely useful
• Help existing customers succeed
• Introduce your category to the right market

If a proposed page cannot explain its job, it should not be hired.

The prescription

1. Choose three to five core commercial topics closely tied to your product and ideal customer.
2. Inspect the SERP for each topic before deciding what page type to create.
3. Build clusters, with one strong primary page and supporting content for distinct related intents.
4. Score keyword opportunities by customer value, intent, feasibility, and product fit—not volume alone.
5. Use customer language from calls, tickets, reviews, and Search Console to find high-quality long-tail opportunities.
6. Prune ruthlessly. If a keyword brings the wrong people, serves no business goal, or requires mediocre content, let it go.

You are not trying to catch every keyword. You are trying to become the obvious answer for the people most likely to care.