Chapter 7: Keyword Research — Finding the Terms People Search (and the Ones You Can Rank For)

"The words are not the point; the person typing them is. Keyword research is the discipline of hearing what your customers actually want — and being honest about which of those conversations you can realistically join." — a constructed epigraph, in the voice of this book

Overview

In Chapter 3 you learned to ask the most important question in SEO — what does the searcher want? — and to read the answer off the results page. This chapter is about the question that comes right after it, and it has two halves that most people only ever ask the first of: what, exactly, do my customers type into Google — and which of those searches can I realistically win?

Notice how different that is from what "keyword research" usually means. Ask a business owner for their keywords and you will get a list of the words they use: "HVAC services," "residential electrical contractor," "plumbing solutions." Those are the words on their van, their invoices, their About page. They are almost never the words their customers type. The customer whose furnace is making a horrible banging noise at 11 p.m. does not search "residential HVAC diagnostic services." They search "furnace making loud banging noise," or "furnace won't turn on," or "emergency furnace repair near me." Keyword research is the discipline of closing that gap — of trading the language of the business for the language of the customer, because the customer's language is what actually gets typed into the box.

And then the second half, the one that separates a professional from an enthusiast: even once you know what people type, you cannot chase all of it. Some searches are contested by national brands with a decade of authority you will not overtake this year. Some are so broad they mean nothing. Some carry no commercial value at all. The art is not finding keywords — a tool can vomit ten thousand of those in a second. The art is finding the specific searches that are valuable to your business and winnable by your site, and spending your finite effort there. This chapter gives you both the free tools to find the words and the honest framework — opportunity as value times winnability — to choose among them.

In this chapter, you will learn to:

• Trade "what words describe my business" for "what does my customer actually type," and generate ideas from seeds, customer language, competitors, and your own Search Console data.
• Run the entire free toolkit — Search Console, Keyword Planner, Google Suggest, People Also Ask, and AnswerThePublic — and know what each is good and bad at.
• Understand what a paid tool (Ahrefs, Semrush) really sells you, and when it is and isn't worth the money.
• Read search volume, difficulty, cost-per-click, and trend for what they are — useful, imperfect estimates — and never mistake a tool's confident number for a fact.
• See why the "long tail" of specific, lower-volume searches is where new and small sites actually win.
• Group keywords into topics, map one primary intent per page, and prioritize the whole list by opportunity.

Learning Paths

Keyword research is foundational for everyone, but the kind of keyword you hunt differs by track. 🏪 Local Business: your gold is service × city and "near me" terms — weight §7.2 (Search Console and Suggest), §7.5 (long-tail local), and the Strategy File, and remember from Chapter 3 that many of these are local-pack fights, not content fights. 📝 Content Creator: informational long-tail is your homeland — live in §7.1 (customer language), §7.5 (the long tail), and §7.6 (grouping into topics). 🛒 E-Commerce: weight §7.4 (CPC as a commercial-value signal) and §7.6 (mapping category vs. product vs. guide keywords); the depth is Chapter 31. 🔧 Developer: §7.6 and §7.7 decide which pages should exist before you build anything — keyword mapping is an architecture input. 📊 Strategist: the whole chapter, especially §7.7's prioritization matrix, which is the engine of every content roadmap you will ever build.

7.1 Seeds: where keyword ideas come from

Every keyword list starts as a handful of obvious terms and grows outward. Those first obvious terms have a name. A seed keyword is a broad, root term — usually one or two words — that describes a core topic of your business and serves as the starting point from which you generate more specific keyword ideas. For Rivertown Home Services, the seeds are almost embarrassingly simple: furnace, air conditioner, water heater, drain, electrical panel, thermostat, HVAC, plumbing, electrician. You do not need cleverness to write the seed list. You need it to know what to do next.

First, though, the term at the center of this whole chapter. A keyword is a word or phrase that people search for, considered as a target you want your content to rank for. It is the same thing Chapter 3 called a query — the literal text typed into the box — seen from the other side of the glass: the searcher types a query; the optimizer targets a keyword. The industry says "keyword" out of habit even when the term is a whole phrase or question ("why is my furnace blowing cold air" is, in SEO jargon, a "keyword"), and even though you now know from Chapter 3 that the word is never really the target — the intent behind it is. Keep that Chapter 3 lesson bolted on: a keyword is a clue to a want, and you are ultimately researching wants. But you research them through the words people actually use, and that is what this chapter equips you to do.

🔗 Connection Keyword research sits downstream of intent, not upstream. Chapter 3 (Search Intent) taught you to classify what searchers want and to read it off the SERP (Search Engine Results Page); this chapter finds the specific words that carry those wants and judges which you can win. When you evaluate a keyword here, you are always also asking Chapter 3's question — "what does someone searching this actually want, and does my page match it?" The two chapters are one motion: understand the want, then find and score its words.

The whole job of §7.1 is expansion: turning a dozen seeds into hundreds of real candidate keywords. The sources, in rough order of value, are these.

Your customers' actual language. This is the richest and most-ignored source, and it costs nothing. The words your customers use are already sitting in your sales calls, your support emails, your booking-form notes, your online reviews, and the mouths of your front-desk staff. People do not call in asking for a "heat exchanger inspection." They say "my furnace smells like burning" and "there's water around the bottom of my water heater" and "half my house lost power." Those phrasings are keywords — the exact ones real people type, because a person describes a problem to Google the same way they describe it to a human. Mine them. A single afternoon reading a year of reviews and support tickets will hand you phrasings no tool would ever surface, because tools reflect what people have searched, and your customers occasionally invent language that has not hit critical mass yet.

Competitors. The businesses already ranking for your topics have, wittingly or not, done keyword research you can read for free. Look at the pages a competitor has built, the headings they use, the questions they answer, the blog topics they cover. You are not copying — you are reading the market's existing map. (Paid tools turn this into a precise "here is every keyword this competitor ranks for that you don't" report; we get there in §7.3 and, properly, in Chapter 30.)

Google itself. Google will hand you an enormous quantity of real keyword ideas through Suggest, People Also Ask, and related searches — all of which reflect genuine searches by real people. This is the free toolkit of §7.2, and it is so central that it gets its own section.

Your own Search Console. If your site has any history at all, Google is already telling you which searches you appear for — including hundreds you never targeted and did not know you ranked for. This is often the single best seed source of all, because it is your real data, not an estimate. Also §7.2.

FROM SEEDS TO A KEYWORD LIST                                    [schematic — not to scale]

   A FEW SEEDS                    EXPANSION SOURCES                 A REAL LIST
   ───────────                    ─────────────────                 ───────────
   furnace          ──┐          • customer language      ──┐      furnace making banging noise
   water heater       ├──▶       • competitor pages          ├──▶  water heater leaking from bottom
   electrical panel   │          • Google Suggest / PAA       │     how much to replace a water heater
   drain              │          • related searches           │     emergency electrician near me
   thermostat       ──┘          • YOUR Search Console       ──┘    furnace repair Cedar Hills
        (9 seeds)                     (free)                          (→ hundreds of candidates)
        │                                                                     │
        └──────── expansion turns a business's vocabulary into ──────────────┘
                  its customers' vocabulary — the words actually typed.

The output of §7.1 is deliberately messy: a big, unfiltered pile of candidate keywords, phrased the way real people phrase things. Do not judge them yet — resist the urge to delete the ones that look low-volume or unwinnable, because you cannot know that until §7.4 and §7.7. The mistake here is premature filtering, throwing away the specific, high-intent long-tail phrases (§7.5) that turn out to be your best opportunities, because they "looked small." Gather first. Judge later.

🔄 Check Your Understanding A plumber writes down his keyword seeds as: "plumbing services," "licensed plumber," "professional drain solutions." What is the core problem with this list, and where would you send him to fix it?

Answer

Every term is the business's language, not the customer's. Real customers search "toilet won't stop running," "drain clogged," "no hot water," "burst pipe emergency." The fix: mine his customers' actual words — reviews, support calls, booking notes — and Google Suggest/PAA, which reflect real searches. He is describing his services; his customers are describing their problems, and Google matches the problem.

7.2 The free toolkit: everything you need to start, at no cost

Here is a claim that saves small businesses thousands of dollars a year: for a first real keyword list, you do not need to pay for anything. Google and a couple of free tools will hand you more genuine, real-world keyword data than a beginner can act on. Paid tools (§7.3) are wonderful and, for professionals, worth every penny — but they are an accelerant, not a prerequisite. Let us walk the free toolkit in order of value.

1. Google Search Console — your own real data. If your site has existed for more than a few weeks, this free Google tool already knows every search query for which your pages have appeared, how many times (impressions), how often people clicked (clicks), and roughly where you ranked (average position). That is not an estimate of what might work — it is a record of what your site is already doing, including, almost always, dozens or hundreds of queries you never deliberately targeted and had no idea you ranked for. There is no better seed source in existence, because it is real, it is yours, and it is free. We set Search Console up properly and read every report in Chapter 27; here we use just one move, and it is the highest-leverage move in this entire chapter.

The move is to find your striking-distance keywords — the queries where you already rank on the edge of page one (roughly positions 8–20) and get impressions but few clicks. These are searches Google already thinks you are almost good enough for. Nudging a page from position 12 to position 6 for a term you already rank for is very often easier, faster, and cheaper than trying to rank a brand-new page for a brand-new term from scratch — because most of the hard work (being indexed, being deemed relevant at all) is already done.

📄 Read the Report

text FIGURE 7.1 — "The goldmine hiding in your own Search Console" [constructed teaching example] THE QUERY / PAGE Rivertown's Search Console Performance report, last 3 months, sorted by impressions, filtered to average position 8–20. WHAT'S THERE "water heater replacement cost" — 1,400 impressions, 11 clicks, avg position 12. "how to relight furnace pilot light" — 900 impressions, 30 clicks, avg position 9. "furnace short cycling" — 600 impressions, 4 clicks, avg position 16. Dozens more the team never targeted, all sitting just off page one. WHAT IT SHOWS Google already considers these pages relevant enough to show — they are on the doorstep of page one. The high-impression, low-click, position-11-ish rows are the ripest fruit on the site: proven demand, proven relevance, a small gap to close. WHAT IT DOESN'T It doesn't tell you WHY each sits where it does (weak title? thin content? stronger competitors?), or guarantee that improving the page will move it — Google decides that. And it only shows terms you ALREADY appear for, not the whole opportunity space. THE MOVE Prioritize improving the pages behind these striking-distance queries (better intent match, comprehensiveness, internal links — Chapters 9 and 15) before chasing brand-new terms. Start where Google has already half-agreed with you. THE LESSON Your own Search Console is the first and best keyword tool you own. Mine what you nearly rank for before you hunt for what you don't rank for at all.

2. Google Keyword Planner — free volume and ideas, with an asterisk. Keyword Planner is Google's own tool, free with any Google Ads account (you do not have to run ads to use it). Feed it a seed and it returns hundreds of related keyword ideas plus an estimate of each one's search volume. Two honest caveats, which most tutorials skip. First, without an active, spending ad campaign, Keyword Planner shows volume in broad ranges or buckets (for example, "1K–10K") rather than precise numbers — Google reserves the granular data for its advertisers. Second, it is built for advertisers, so it groups closely related terms together and leans toward commercially valuable keywords. It is genuinely useful for idea generation and rough volume, and it has the authority of being Google's own data — just do not treat its numbers as exact.

3. Google Suggest (autocomplete) — real demand, live. Start typing a seed into Google and watch the dropdown fill with completions: type "furnace " and you get "furnace filter," "furnace not turning on," "furnace making noise," "furnace repair near me." Those suggestions are generated from real, aggregated searches — Google is literally showing you things people commonly search. It is one of the purest free windows into genuine demand that exists. The professional's trick is to work the alphabet: type your seed followed by each letter ("furnace a," "furnace b," "furnace c" …) and by question words ("why is my furnace," "how to furnace," "furnace vs …"), harvesting the suggestions at each step. In twenty minutes you will have a page of real, human phrasings no business would have invented on its own.

4. People Also Ask — the question mine. On a great many results pages, Google shows an expandable box of related questions under the heading "People also ask." People Also Ask (PAA) is that on-SERP feature: a set of related questions Google displays, each of which expands to a brief answer and — the useful part — often spawns more questions when you click it, so the box regenerates almost endlessly. For "water heater," PAA might offer "How long do water heaters last?", "How much does it cost to replace a water heater?", "What are the signs a water heater is going bad?" Each is a real question real people ask, which means each is a candidate keyword and a sub-topic your page may need to cover to be comprehensive. PAA is thus doing double duty: it is a keyword source here, and — as you will see in Chapter 10 — a structural signal about what a complete answer looks like.

5. AnswerThePublic and friends — Suggest, visualized. A category of free-tier tools (AnswerThePublic is the best known) does the alphabet-soup Suggest harvest for you and arranges the results into tidy wheels of questions, prepositions, and comparisons ("water heater for," "water heater vs," "water heater without"). These tools are mostly repackaging Google Suggest data, so they are not a new source so much as a faster way to harvest an old one. The free tiers limit how many searches you can run per day, which is fine for a beginner doing occasional research. Useful, not magical — and nothing they show you that you could not get, slower, by hand.

🔎 How Search Sees It Suggest, PAA, and related searches are windows into Google's aggregated record of what real people actually type — which is exactly why they are trustworthy as demand signals and a little dangerous as volume signals. They tell you, reliably, that a phrasing is real and common (Google would not surface it otherwise). They do not tell you how common, and Google filters and moderates these features (it suppresses some offensive, dangerous, or personal suggestions, and it personalizes them somewhat). So read them as a true map of the territory of demand — these are genuine roads people travel — while getting your rough volumes from Keyword Planner and your own Search Console. Shape from one source, size from another.

🛠️ Try It on Your Site Do the two free moves that matter most, right now, in fifteen minutes. (a) Harvest Google Suggest. Pick one seed for your business. In an incognito window, type it into Google followed by a space and each letter a–z, then by "how," "why," "best," "vs," and "near me," writing down every autocomplete suggestion and every "People also ask" question you see. You will end with 30–50 real phrasings your customers actually use. (b) Mine your Search Console. Open the Performance report (or, if you have no Search Console yet, this is your reason to set it up — Chapter 27). Sort by impressions, then look for queries at average position 8–20 with lots of impressions and few clicks. Those are your striking-distance keywords: searches you nearly rank for. Write down the top ten. You have just found, for free, both what people type (a) and where you are closest to winning (b) — the two things this whole chapter is about.

7.3 The paid toolkit: what you're really buying

At some point — usually when keyword research stops being a one-time task and becomes a recurring job — the free toolkit's friction starts to cost more than a subscription would. That is when a paid tool earns its place. The major platforms (Ahrefs, Semrush, and Moz are the best known; there are others) cost somewhere in the range of a modest monthly software subscription up to a serious one, and it is worth being clear-eyed about what that money actually buys, because the marketing around these tools implies it buys rankings. It does not. It buys information and speed.

Here is what you are really paying for:

• Scale. A vast keyword database you can query in seconds — enter a seed and get thousands of related terms with volume, difficulty, and more, instantly, instead of harvesting Suggest by hand.
• Competitor intelligence. The genuine superpower: enter a competitor's domain and see (an estimate of) every keyword they rank for, which of their pages earn the most traffic, and — most valuable of all — the keyword gap: terms your competitors rank for that you don't. This is the paid toolkit's killer feature, and it is the heart of the competitive analysis we cover in Chapter 30.
• Difficulty scores at scale. A proprietary "how hard is this to rank for" number for every keyword (dissected honestly in §7.4), so you can filter thousands of terms quickly — a rough first pass, not a verdict.
• Backlink data. Who links to whom, which feeds both difficulty estimates and the link-building work of Part IV (Chapters 22–26).
• Rank tracking and monitoring. Automated tracking of where you rank over time for your priority terms (Chapter 30).

And here, in the same breath, is what that money does not buy, because a tactic without its limits has failed:

• Not Google's data. Every volume and difficulty number in every paid tool is an estimate built from the tool's own crawling, clickstream data, and modeling. Different tools give different numbers for the same keyword — sometimes wildly different — because they are all guessing at data only Google truly has. A paid number is not more true than a free one; it is more convenient.
• Not a strategy. A tool will happily generate a list of ten thousand keywords. It will not tell you which matter to your business, which match an intent you can serve, or which are worth your finite effort. That judgment — the actual work — is §7.6 and §7.7, and no subscription performs it for you.
• Not a requirement to start. A local business building its first keyword list does not need one. The free toolkit plus reading SERPs by hand gets you most of the way; the paid tool mainly saves time, which matters enormously once your time is the scarce resource, and not at all when it isn't.

🚫 SEO Myth: "You need an expensive tool to do keyword research." This is a belief the tool vendors are, understandably, in no hurry to correct, and it intimidates beginners into thinking real SEO starts at \$100+ a month. It doesn't. Everything in §7.2 is free, and for a single small business the free toolkit plus your own Search Console plus reading the SERP yourself produces a keyword list you can act on for a year. Paid tools are genuinely excellent and, for anyone doing SEO professionally or at scale, well worth the cost — their competitor and gap analysis alone can repay the subscription. But "I can't do keyword research until I buy Ahrefs" is false, and it has stopped more beginners than it should. Start free. Upgrade when your time, not your ambition, demands it.

🔗 Connection The paid tools' standout feature — seeing what competitors rank for and finding the keyword gap and content gap between you and them — is the subject of Chapter 30 (Rank Tracking and Competitive Analysis), and it feeds the comprehensive SEO audit of Chapter 38. Here we only note that this is what you are buying; there we put it to work.

7.4 The metrics: volume, difficulty, CPC, and trend (and what each hides)

Every keyword tool decorates each term with numbers, and those numbers are how you turn a giant undifferentiated list into a ranked one. Four metrics carry most of the weight. Learn what each is, and — because this is a book about honesty in a field full of confident fakery — learn what each one quietly hides.

Search volume. Search volume is the estimated number of times a keyword is searched in a given period, almost always expressed as a monthly average. It is the closest thing to a "how much demand is there" number, and it is the first thing everyone looks at. It is also, of the four, the one most often misread as precise when it is nothing of the kind.

⚖️ Evidence Check Claim: "This keyword gets 2,900 searches a month." How much should you trust that number? — What's real: the relative magnitude. A keyword with an estimated 5,000 monthly searches genuinely has far more demand than one estimated at 50. Volume estimates are reliable enough to rank and compare keywords, which is most of what you need. — What's not: the precision. That "2,900" is an estimate — Google's own Keyword Planner reports volume in broad buckets for non-advertisers, third-party tools model it from samples, and the true figure swings with seasonality (a "furnace repair" search spikes in the first cold snap and cratered in July), news, and trends. Different tools routinely report different volumes for the same term. So use volume as a rough size band — "big / medium / small / tiny" — not as a number you would put in a contract. Anyone who forecasts traffic to three significant figures from a keyword-tool volume is selling false precision, the exact sin Chapter 2 warned you to distrust.

Keyword difficulty. Keyword difficulty is a score — usually on a 0–100 scale — that estimates how hard it will be to rank on page one for a keyword. It is the metric that most tempts misuse, because it looks like an objective measurement and is nothing of the sort.

🚫 SEO Myth: "Keyword difficulty is an objective number from Google." It is neither objective nor from Google. Google publishes no difficulty score. Every "KD" number you see is a proprietary estimate invented by a tool vendor, and each vendor computes it differently — mostly from the backlink profiles (Chapter 22) of the pages currently ranking, sometimes with other factors mixed in. That is why the same keyword can score 34 in one tool and 51 in another: they are different guesses, not readings of one true value. Worse, a difficulty score is generic — it estimates difficulty for an average site, not for your site with your authority in your niche. A term that is "hard" for a new blog may be easy for an established local brand, and vice versa. Use difficulty scores as a coarse first filter to sort thousands of terms fast — but the real difficulty read comes from opening the SERP and looking at who actually ranks (§7.7). The score narrows the field; your eyes make the call.

Cost-per-click (CPC). CPC (cost-per-click) is the average amount advertisers pay Google for one click on a paid ad for that keyword. You might wonder why an advertising number matters to organic SEO. It matters because it is the best free proxy you have for a keyword's commercial value. Advertisers bid real money on clicks, and they bid the most on searches by people about to spend money. So a high CPC is a strong signal that a keyword is commercially valuable — that its searchers convert into customers. "Water heater installation cost" carries a meaningful CPC because those searchers are near a purchase; "how does a water heater work" carries almost none, because those searchers are just learning. Read alongside intent (Chapter 3), CPC helps you tell the money keywords from the merely-popular ones.

⚖️ Evidence Check Claim: "High CPC means a keyword is worth targeting; low CPC means it isn't." The honest version is more nuanced. — What CPC reliably signals: commercial intent. Advertisers pay more for clicks that turn into revenue, so a high CPC is genuine evidence that a keyword's searchers are close to buying. That much is sound. — What it doesn't mean: that low-CPC keywords are worthless. A low-CPC informational query ("why is my furnace short cycling") may bring the visitor who, an hour later, books a repair — value that shows up as a later conversion, not an immediate one, and that CPC cannot see. CPC measures advertiser competition for an immediate click, which correlates with but is not identical to value to your business over the whole customer journey. Use CPC to spot the money terms; do not use it to dismiss the top-of-funnel terms that feed them.

Trend. A keyword's volume is a snapshot; its trend is the direction that snapshot is moving. Google Trends (free) shows a term's relative interest over time — rising, falling, flat, or seasonal. Trend matters for two reasons. First, seasonality: "furnace repair" and "air conditioner installation" are violently seasonal, and an annual-average volume hides the peaks you actually need to prepare content for (you want the furnace guide ranking before the first freeze, not in response to it). Second, trajectory: a term with modest but rising volume ("heat pump water heater") may be a better long-term bet than a bigger but declining one, because SEO is a long game (theme 6) and you are investing in an asset that will still be relevant in two years. Trend is how you avoid pouring effort into a fading topic or missing a rising one.

Here is the whole quartet in one place — including, in the last column, the caveat you must carry for each.

A note on the light arithmetic you will be tempted to do. It is reasonable to estimate a keyword's rough traffic potential — very loosely, its search volume multiplied by the share of clicks a good ranking earns. Suppose (all numbers illustrative) a term gets about 1,000 searches a month and you believe reaching a strong position might earn on the order of ten percent of those clicks: that is roughly 1,000 × 0.10 = 100 visits a month, before you write a word. Treat that as a fuzzy order-of-magnitude sketch, never a forecast — the click-through rate by position is itself a curve we do not pin to fake-precise numbers until Chapter 39, where the forecasting math gets its careful, honest treatment. For now: volume gives you size, and a rough click share turns size into a sense of traffic. A sense is enough to prioritize; it is not enough to promise.

7.5 Long-tail keywords: lower volume, higher intent, easier wins

Beginners chase the big, short, obvious keywords — "plumber," "water heater," "HVAC" — the ones with the gaudiest volume numbers. It is almost always the wrong instinct, and understanding why is the difference between a keyword strategy that works for a small site and one that never gets off the ground. The answer lives in the long tail.

A long-tail keyword is a longer, more specific search phrase with lower individual search volume, usually carrying clearer intent and facing less competition than a short, broad "head" term. "Water heater" is a head term: two words, enormous volume, impossibly broad intent (learn? buy? repair? which kind?), and contested by every retailer, manufacturer, and publisher on Earth. "40-gallon gas water heater keeps running out of hot water" is long-tail: specific, lower-volume, unmistakable intent (a person with a specific problem), and contested by far fewer, weaker pages. The name comes from the shape of search demand when you plot it.

THE SEARCH DEMAND CURVE — WHY THE TAIL WINS                    [schematic — not to scale]

  search
  volume
   │
   │ █                          "water heater"  ← the HEAD:
   │ █                            a few huge, broad, brutally
   │ █ █                          contested terms
   │ █ █ █                      "water heater replacement cost"  ← the BODY:
   │ █ █ █ █                      more specific, moderate volume
   │ █ █ █ █ █ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄ ▄  "40-gal gas water heater not enough hot water"
   │ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █ █  ← the LONG TAIL:
   └───────────────────────────────────────────────────────    thousands of specific,
        few terms  ───────────▶  very many terms                low-volume, high-intent,
                                                                low-competition searches

   The head is tall but narrow. The tail is short but ENDLESS — and, added up across
   thousands of specific searches, the tail's total volume typically DWARFS the head's.

Read that curve and the strategy falls out of it. The head terms are a handful of tall bars: huge volume, but few in number and defended by the strongest sites on the web. The long tail is thousands upon thousands of short bars: each tiny on its own, but so numerous that their sum usually exceeds the head entirely. And the tail has three properties that make it where new and small sites actually win:

• Clearer intent, so higher conversion. "Water heater" could mean anything; "emergency water heater repair Northgate" means one very specific, very ready-to-act thing. A visitor from a specific long-tail search is far likelier to become a customer than a visitor from a vague head term, so long-tail traffic converts better even though there is less of it.
• Less competition, so winnable. Fewer, weaker pages target any given long-tail phrase, because most sites chase the head. A new site has a real chance at "furnace making a clicking noise then shuts off" and no chance at "furnace." Winnability (§7.7) is dramatically higher down the tail.
• Compounding coverage. You cannot rank for the head until you have earned authority — and one of the best ways to earn it is to become genuinely comprehensive across the tail. Owning a thousand specific answers about water heaters is how a site eventually becomes the kind of authority that can compete for "water heater" itself. The tail is not a consolation prize; it is the on-ramp.

There is an honest limit, as always. Long-tail keywords are individually low-volume — some so rare that a single one is not worth a dedicated page. The long-tail strategy is a strategy of aggregation: you win by covering many related specific searches well, usually grouped onto comprehensive pages (§7.6), not by building a thin page for every three-word variation. And "long-tail" is not a magic wand that removes all competition — some specific phrases are contested precisely because they are valuable. The tail is where the winnable opportunity concentrates; it is not a guarantee, and it still requires the prioritization of §7.7.

🔗 Connection For a business like Rivertown, the richest long tail is service × city: "furnace repair Cedar Hills," "drain cleaning Northgate," "panel upgrade Westbrook." Building those pages well — genuinely local and distinct, not thin near-duplicates — is exactly the line between a legitimate long-tail strategy and doorway-page spam, which is why the service×city page system gets its careful treatment in Chapter 33 (Programmatic and Enterprise SEO) and the local dimension in Chapter 25 (Local SEO). Here we only note that the tail is where these pages live; there we build them without crossing the line.

🔄 Check Your Understanding A new home-services site with little authority has two candidate keywords: "water heater" (volume ~90,000, difficulty 80) and "water heater making popping noise" (volume ~500, difficulty 15). Which should it target first, and why — and what happens to the head term over time?

Answer

Target the long-tail "water heater making popping noise" first. It is winnable (low difficulty, weak competition), carries clear intent (a person with a specific problem, likely to need a pro), and converts better despite lower volume. "Water heater" is unwinnable for a new site today — but by covering hundreds of specific water-heater searches well, the site builds the topical authority (Chapter 4) that eventually makes the head term reachable. The tail is the on-ramp to the head, not a substitute for it.

7.6 Grouping keywords into topics: one page, one primary intent

By now your list has hundreds of keywords, and a dangerous temptation has arrived with them: to build one page per keyword. Resist it completely. Many of your keywords are not different topics at all — they are the same question wearing different clothes, and they want one page, not ten.

Consider these five keywords: "water heater replacement cost," "how much to replace a water heater," "cost to install a new water heater," "water heater replacement price," "average cost of new water heater." A beginner sees five keywords and imagines five pages. But these are not five wants; they are one want — "what will it cost me to replace my water heater" — phrased five ways. They share a single intent, and one genuinely good page should target all five. Building five near-identical pages instead does active harm: they compete with each other for the same ranking (a problem called keyword cannibalization, which is Chapter 12's territory), they split your effort and authority, and they clutter your site with thin duplicates that quality systems (Chapter 6) dislike. One want, one page.

So after gathering (§7.1–7.2) and scoring (§7.4), the next move is grouping: clustering your keyword list into sets that share a single primary intent, where each set will become one page. The most reliable test for whether two keywords belong together is not how similar the words look — it is whether they share a SERP.

🔎 How Search Sees It Here is a free, decisive test for whether two keywords are really one topic: search both and compare the results. If Google returns largely the same pages for "water heater replacement cost" and "how much to replace a water heater," then Google considers them the same underlying need — and one page can and should target both. If the two queries return different sets of pages, Google sees two distinct intents, and they want two different pages. This "same-SERP test" is more trustworthy than any keyword-similarity score or your own intuition, because it reads the intent directly off Google's own behavior — the Chapter 3 answer-key move, applied to grouping. When in doubt about whether to split or combine, let the SERPs decide.

Once your keywords are grouped into intent-sharing topics, you assign each topic to a page. That assignment has a name. Keyword mapping is the practice of assigning each keyword (or group of keywords) to the single, specific page that will target it — one primary intent per page — so that every important keyword has a clear home and no two pages fight over the same one. The deliverable is a simple map: this group of keywords → this page; that group → that page. It is unglamorous and it is the backbone of a coherent site, because it turns a pile of keywords into a plan for pages, and it guarantees that when you sit down to write, each page knows exactly which want it exists to satisfy.

GROUPING AND MAPPING                                           [schematic]

  A PILE OF KEYWORDS                GROUPED BY SHARED INTENT          MAPPED TO PAGES
  ─────────────────                ────────────────────────          ───────────────
  water heater replacement cost ┐
  how much to replace a heater  ├─▶ COST / DECISION topic     ──────▶ /water-heater-replacement
  cost to install water heater  ┘   (one want, five phrasings)         (one page, one intent)

  how to relight pilot light    ┐
  furnace pilot light won't stay├─▶ PILOT-LIGHT HOW-TO topic  ──────▶ /blog/relight-furnace-pilot
  relight furnace pilot         ┘                                     (one page, one intent)

  furnace repair Cedar Hills    ────▶ LOCAL SERVICE (per city) ─────▶ /furnace-repair/cedar-hills
                                                                       (one page per city)

This grouping-and-mapping step is also the natural bridge out of keyword research and into content strategy. A mapped keyword list is not yet a content plan — it does not yet say which pages are "pillars" and which are supporting "clusters," in what order to build them, or how they interlink — but it is the raw material from which that plan is built.

🔗 Connection Grouping keywords into intent-topics is the doorstep of topic clusters and the pillar-and-cluster model, which are the heart of Chapter 8 (Content Strategy). The idea that a site earns topical authority by comprehensively covering a subject's cluster of related terms is Chapter 4's. The danger of two pages targeting one keyword — keyword cannibalization — is diagnosed and fixed in Chapter 12. This section produces the mapped list; those chapters turn it into an architecture and a calendar. Map here; build there.

7.7 Prioritization: the opportunity = value × winnability matrix

You now have a grouped, mapped, metric-decorated keyword list — and far more opportunities than you have time to pursue. Prioritization is the last and most important skill in the chapter, because a keyword list you cannot rank against your available effort is just an inventory of frustration. The professional's tool for it is a single, honest idea:

Opportunity = Value × Winnability.

A keyword is worth pursuing to the degree that it is both valuable (ranking for it would meaningfully help the business) and winnable (you can realistically reach page one for it with the effort and authority you have). Both factors matter, and multiplying them — rather than adding — captures the truth that a keyword weak on either dimension is a poor target, no matter how strong it is on the other. The biggest, most valuable term in your industry is worthless to you if you cannot rank for it; the most winnable term on Earth is worthless if ranking for it brings no value.

Unpack the two factors:

Value is not just search volume. It is volume weighted by intent and commercial worth — the Chapter 3 and §7.4 lessons, combined. A low-volume transactional term ("emergency furnace repair Millhaven") can be worth far more to Rivertown than a high-volume informational one ("what is a BTU"), because its searchers are ready to hire. Value = how much would ranking here actually move the business? — which folds together volume, intent, CPC/commercial value, and how directly the searcher connects to revenue.

Winnability is your realistic chance of ranking, and here is the section's central practical claim: you judge winnability by reading the SERP, not by trusting a difficulty score. A generic difficulty number (§7.4) is a coarse first filter. The real read comes from opening the results page and looking hard at who is actually there.

📄 Read the SERP

text FIGURE 7.2 — "Judging winnability by reading the SERP" [constructed teaching example] THE QUERY / PAGE Two candidate keywords for Rivertown, each SERP read by eye. WHAT'S THERE "water heater" — page one is Home Depot, Lowe's, Rheem, This Old House, Wirecutter: national retailers, manufacturers, and major publishers, every result a heavyweight. "furnace making a rattling noise" — page one is two forum threads, a couple of small HVAC-contractor blogs, a Q&A site, and one mid-size home site. No national brands. WHAT IT SHOWS "water heater" is effectively unwinnable for a small local site — you would be the smallest fish in an ocean of whales. "furnace making a rattling noise" is genuinely winnable: the incumbents are weak (forums, thin blogs), and a real HVAC company writing a clear, expert answer could plausibly outrank them. WHAT IT DOESN'T A weak-looking SERP is an *opportunity signal*, not a promise — ranking still requires a genuinely better page, and Google decides. And a difficulty *score* might have rated both terms without capturing this vivid difference in who actually ranks. THE MOVE Skip the head term for now (revisit it years later, from authority). Target the winnable long-tail term with a genuinely expert page. Read the SERP for every serious candidate before committing effort. THE LESSON The results page is the truest difficulty meter you own. Weak incumbents = your opening; a wall of national brands = pick a different fight.

What are you looking for when you read a SERP for winnability? Weak incumbents you could plausibly beat: forum threads, thin or outdated pages, results that don't quite match the intent, small sites with no particular authority. Those are openings. And warning signs that you're outmatched: a page full of national brands and established authorities, or — critically for a local business — a local pack, which (as Chapter 3 taught) means the fight is not a content fight at all but a local-SEO fight decided by Google Business Profile, proximity, and reviews (Chapter 25). Reading the SERP tells you not just how hard but what kind of fight you are in.

Put value and winnability on two axes and the whole list sorts itself into four quadrants:

THE OPPORTUNITY MATRIX — VALUE × WINNABILITY                   [schematic]

            HIGH WINNABILITY                 LOW WINNABILITY
          (you can realistically         (national brands, huge
           reach page one)                authority, or a local pack)
        ┌──────────────────────────┬──────────────────────────────┐
  HIGH  │  ★ DO THESE FIRST         │  ⏳ THE LONG GAME             │
 VALUE  │  High-intent, winnable    │  Valuable but contested.     │
        │  long-tail & service×city │  Build toward these over     │
        │  terms. Your bread and    │  months/years as authority   │
        │  butter — start here.     │  grows. Don't start here.    │
        ├──────────────────────────┼──────────────────────────────┤
  LOW   │  ✔ QUICK WINS / FILL      │  ✖ SKIP                      │
 VALUE  │  Easy but low-value.      │  Hard AND low-value.         │
        │  Useful for coverage &    │  No reason to spend effort   │
        │  authority; do when cheap.│  here. Ignore.               │
        └──────────────────────────┴──────────────────────────────┘

The matrix turns an overwhelming list into a sequence. Start in the top-left — high value, high winnability — because those are the keywords that both matter and you can actually reach: for Rivertown, the specific service × city and high-intent long-tail terms. Treat the top-right (high value, low winnability) as the long game — the head terms and fiercely contested queries you build toward over months and years as your authority grows, exactly the compounding, patient work theme 6 insists SEO really is. Use the bottom-left (easy but low-value) as filler and coverage that quietly builds topical authority when it's cheap to produce. And ignore the bottom-right entirely — hard and low-value is nobody's good use of a finite week.

⚖️ Evidence Check Claim: "This keyword has a difficulty of 22, so I can definitely rank for it." Where's the honesty line? — Fair to say: a low difficulty score, confirmed by reading a weak SERP, means a term is plausibly winnable — the incumbents look beatable and the odds are reasonable. That is a sound basis for prioritizing effort. — Not fair to say: "definitely," or any promise of a ranking. Difficulty scores are generic vendor estimates (§7.4), your real odds depend on your site and the effort you invest, and Google decides the ranking, not the tool and not you. Winnability is a considered bet that improves your odds — never a guarantee. The whole book refuses the word "guaranteed," and prioritization is no exception: you are choosing the fights most likely to reward the work, not fights you are certain to win.

📈 The Strategy File

Time to put the chapter to work on Rivertown Home Services. Chapter 3 gave us the intent of Rivertown's core queries (Figure 3.7); this chapter's job is to turn that into a starter keyword list — a real, prioritized inventory of service × city and informational terms — scored by opportunity (value × winnability). This is the raw material every later content decision will draw on. (All Rivertown volumes, difficulties, and scores below are a constructed teaching example — illustrative bands, not real measurements.)

We build it in three moves, exactly as the chapter taught. Expand the seeds (furnace, AC, water heater, drain, panel, thermostat) across the five cities — Rivertown, Cedar Hills, Northgate, Westbrook, Millhaven — and across the informational questions customers actually ask (harvested from Suggest, PAA, and Rivertown's own reviews and Search Console). Score each with rough value and winnability. Prioritize with the matrix.

FIGURE 7.3 — "Rivertown's starter keyword list, scored by opportunity"    [the Strategy File]
  KEYWORD (example)              INTENT (Ch 3)     VALUE   WINNABILITY        QUADRANT / MOVE
  furnace repair Cedar Hills     transactional/    HIGH    MED-HIGH (local    ★ DO FIRST — service×city
                                  local             (ready   pack + a real,     page, built genuinely local
                                                    to hire)  distinct page)     (Ch 15/25/33)
  emergency electrician          transactional/    HIGH    MED (local-pack    ★ DO FIRST — but largely a
    Northgate                     local             (urgent) fight; content     local-pack fight (Ch 25),
                                                             alone won't win)   not a content one
  water heater replacement       commercial/       HIGH    MED (striking-      ★ DO FIRST — reshape the #11
    cost                          informational      (near   distance in GSC,   page to serve the decision
                                                     buying)  Figure 7.1)        (intent work, Ch 3/9)
  how to relight a furnace       informational      MED     HIGH (weak SERP:   ✔ QUICK WIN — expert how-to
    pilot light                    (Know)            (feeds   forums, thin       (+ video, Ch 11); builds
                                                     calls)   blogs)             trust & topical authority
  furnace making a rattling      informational      MED     HIGH (weak         ✔ QUICK WIN — genuinely
    noise                          (Know)            (leads   incumbents)        expert answer; on-ramp to
                                                     to jobs)                    authority
  ac repair near me              transactional/     HIGH    LOW-MED (fierce    ⏳ LONG GAME + local pack —
                                  local              (money)  local pack)        win via GBP/reviews (Ch 25),
                                                                                 not a blog post
  water heater                   mixed/broad        HIGH    LOW (Home Depot,   ⏳ LONG GAME — unwinnable
                                                     (huge)   Rheem, Wirecutter) now; build toward it via the
                                                                                 whole tail over years
  what is a BTU                  informational      LOW     HIGH (easy)        ✖ SKIP (for now) — winnable
                                    (Know)            (no      but low value; no path to a
                                                     revenue) customer

Read the figure as a plan taking shape. The top-left of the matrix — Rivertown's real starting work — is the service × city terms and the striking-distance cost page: high value because those searchers are ready to hire or decide, and winnable because they are specific, local, and (for the cost page) already on the doorstep of page one in Search Console. The informational how-tos ("relight a pilot light," "rattling noise") are quick wins: modest direct value, but genuinely winnable against weak SERPs, and they do the quiet compounding work of building trust and topical authority (and catching the DIY searcher who gives up and calls). The head terms and hottest local queries ("water heater," "ac repair near me") are the long game — either genuinely unwinnable today ("water heater") or a local-pack fight that content alone cannot win ("ac repair near me"), both to be built toward rather than chased now. And "what is a BTU" gets skipped: winnable, yes, but it brings no customer.

What this settles, and what it doesn't. It settles Rivertown's priorities — which specific keywords to pursue first, which to build toward, and which to ignore — and it hands the next chapters a real, scored list to work from. It does not settle the content plan (which pages are pillars, which are clusters, in what order, how they interlink — that is Chapter 8), the page architecture (Chapter 15), the local-pack playbook for the "near me" and city terms (Chapter 25), the service×city page system at scale without doorway spam (Chapter 33), or the traffic and ROI forecast those volumes might justify (Chapter 39). And every number here is illustrative — Rivertown's real first task is to replace these bands with actual Search Console and free-tool data using §7.2. Keyword research produces the ranked list of fights worth having. Choosing how to win each one is the rest of the book.

Conclusion

We set out to answer two questions, not one: what do my customers actually type, and which of those searches can I realistically win? The first question retires the amateur's habit of listing the words a business uses about itself and replaces it with the discipline of hearing the words customers use about their problems — mined from customer language, competitors, and, above all, Google's own free tools and your own Search Console. The second question is the one that separates strategy from wishful thinking: you cannot chase every keyword, so you score each by opportunity = value × winnability, judge winnability by reading the SERP rather than trusting a difficulty score, and spend your finite effort where value and winnability meet.

Along the way we stayed honest in the way this book demands. Search volumes are estimates, not measurements. Keyword difficulty is a vendor's guess, not Google's ruling. CPC signals commercial value but misses later-journey worth. The long tail — specific, lower-volume, higher-intent — is where new and small sites actually win, and its total demand dwarfs the head. And no score, however low, promises a ranking; winnability is a considered bet, never a guarantee.

What you have at the end of this chapter is a grouped, mapped, prioritized keyword list — a ranked inventory of the fights worth having. What you do not yet have is a plan for the pages that will win them: which topics become comprehensive "pillar" pages, which become supporting "cluster" content, in what order you build them, and how they link together into a site that earns topical authority. That is the leap from a keyword list to a content strategy, and it is exactly where we go next.

→ Continue to Chapter 8: Content Strategy.

Key Terms

• Keyword — a word or phrase that people search for, considered as a target you want your content to rank for; the same thing Chapter 3 calls a query, seen from the optimizer's side.
• Seed keyword — a broad, root term (usually one or two words) describing a core topic, used as the starting point from which more specific keyword ideas are generated.
• Search volume — the estimated number of times a keyword is searched in a given period (usually a monthly average); always an estimate, often bucketed, and hiding seasonality within its average.
• Keyword difficulty — a proprietary 0–100 score, invented by tool vendors (not Google), estimating how hard it is to rank on page one for a keyword; a coarse, generic first filter, not a reading of your real odds.
• CPC (cost-per-click) — the average amount advertisers pay for one click on a paid ad for a keyword; used in SEO as a free proxy for a keyword's commercial value.
• Long-tail keyword — a longer, more specific, lower-volume search phrase, usually with clearer intent and less competition than a short, broad "head" term; where new and small sites most often win.
• People Also Ask (PAA) — the expandable box of related questions Google shows on many results pages; a free source of real sub-questions, keyword ideas, and comprehensiveness cues.
• Keyword mapping — assigning each keyword (or group of keywords sharing one intent) to the single page that will target it — one primary intent per page — so every important keyword has a clear home and no two pages compete for the same one.

Spaced Review

Retrieval practice mixing this chapter with Chapter 3. Try each before revealing the answer.

1. State the two questions keyword research really answers, and explain why listing "the words that describe my business" fails at both.
2. Name three tools in the free keyword toolkit, and say what "striking-distance" keywords are and why they are the ripest opportunity in your own Search Console.
3. Write the opportunity formula, define its two factors, and explain why you judge one of them by reading the SERP rather than by a difficulty score.
4. (From Chapter 3.) How does the "same-SERP test" for grouping keywords reuse Chapter 3's idea that "the SERP is the answer key"?
5. (From Chapter 3.) A keyword tool labels "ac repair near me" as high-volume and high-value, so a beginner plans a long blog post to rank for it. Using intent (Chapter 3) and winnability (this chapter), explain why that plan is doomed and what the real fight is.