项目文件夹

文件
wehub-resource-sync e011bb26cd
Lint skills / Validate catalog structure (push) Has been cancelled
chore: import upstream snapshot with attribution
2026-07-13 12:36:55 +08:00

11 KiB

Intent Classification Guide

Detailed examples of each of the four search intent categories. Use this to tag every keyword in research with its intent before deciding what content (if any) to build.

Intent classification is the most important and most-skipped step in keyword research. Tools surface volume and difficulty. They do not tell you whether the searcher wants to learn, compare, navigate, or buy. Without intent, you produce content that ranks for words but converts nothing.


The four intents

Intent The searcher wants to Right content Right CTA
Informational Learn something Article, guide, video Subscribe, related content
Navigational Reach a specific destination Branded page, login, location Direct link
Commercial investigation Compare options before buying Comparison, review, list, buyer's guide Demo, trial, deeper guide
Transactional Take an action (buy, sign up, download) Product page, pricing, signup, checkout Buy / sign up / download

The same root word can fall into multiple intents depending on modifier. "Crm" alone is ambiguous. "What is a crm" is informational. "Salesforce login" is navigational. "Best crm for startups" is commercial investigation. "Buy salesforce" is transactional.


Informational intent

The searcher wants to understand, learn, or solve a problem.

Signals

  • Question words: what, why, how, when, where, who
  • "Guide," "tutorial," "examples," "definition," "explained"
  • Topic-only queries with no commercial modifier ("dmarc records," "css grid")
  • "How to" + a task

Examples

Query Intent depth Notes
"what is a service blueprint" Top-of-funnel learning Definition piece
"how to write a creative brief" How-to Step-by-step guide
"service blueprint examples" Mid-funnel learning Examples gallery
"service blueprint vs customer journey map" Comparison-flavored learning Overlaps with commercial; treat as info if no products in result
"why is css grid useful" Conceptual Explainer with code examples
"history of bauhaus design" Background Editorial / long-form
"node js error eaddrinuse" Troubleshooting Specific fix article

What to publish

Long-form articles, guides, glossary entries, video tutorials, original research. Content that earns subscribers, social shares, and backlinks. Not where conversion happens; where awareness builds.

CTA approach

Soft. "Subscribe to the newsletter," "read the related guide," "join the community." A pricing CTA at the end of an informational piece reads as a bait-and-switch.

Failure mode

Producing informational content with hard sales CTAs because someone metric'd it as "low conversion." Informational content's job is awareness and authority. Measure it on assists, brand recall, and downstream conversions, not last-click.


Navigational intent

The searcher knows what they want and is using search to get there.

Signals

  • Brand names ("HubSpot," "Notion")
  • Brand + modifier ("HubSpot login," "Notion pricing," "Salesforce careers")
  • Product names ("MacBook Pro M4," "Tesla Model Y")
  • Domain-style queries ("github.com," "facebook")

Examples

Query What the searcher wants
"linear" Linear's homepage
"linear login" Linear's login page
"salesforce careers" Salesforce's careers page
"amazon prime" Amazon Prime's page
"github status" GitHub's status page

What to publish

For your own brand: own the navigational queries. The homepage, login page, status page, careers page, contact page, pricing page should rank #1 for branded variants.

For other brands' navigational queries: do not pursue. You will not rank, you should not try, and even if you did, the user does not want you.

The exception: comparison queries ("Linear alternatives," "vs Linear") are not navigational; they are commercial investigation. See below.

CTA approach

Whatever the destination expects. Login page → login form. Pricing page → buy or trial. Careers page → application.

Failure mode

Treating "alternatives to" queries as navigational. They are not. They are commercial investigation, and a comparison page is the right format.


Commercial investigation intent

The searcher is in the buying process but not ready to buy. They are comparing, reading reviews, building a shortlist.

Signals

  • "Best [category]" ("best crm for small business")
  • "Top [category]" ("top hosting providers")
  • "[Category] comparison" ("crm comparison")
  • "[Brand] vs [brand]" ("notion vs obsidian")
  • "[Brand] alternatives" ("alternatives to mailchimp")
  • "[Brand] review" ("notion review")
  • "[Category] for [audience]" ("crm for startups")
  • "Cheap / affordable / free [category]"
  • "Buyer's guide for [category]"

Examples

Query What the searcher wants
"best email marketing tool" A ranked list of options
"mailchimp vs convertkit" A direct comparison
"alternatives to mailchimp" A shortlist of competitors with pros and cons
"convertkit review" Honest assessment from a third party (or the brand's own page if positioning)
"crm for solo consultants" Curated shortlist for a niche
"free project management software" Filtered list with feature notes

What to publish

  • For your category (high traffic, broad): "best of" lists, ranked comparisons, buyer's guides. Honest about tradeoffs. Include yourself if relevant; do not lie about ranking.
  • For your brand vs others: comparison pages on your own domain ("[YourBrand] vs [Competitor]"). These are the highest-converting content type after pricing pages.
  • For your category by use case: segment-specific landing pages ("CRM for solo consultants," "CRM for series-A startups"). Use case + CTA.

CTA approach

Stronger than informational, lighter than transactional. "Try [product] free," "see how it compares," "book a demo," "download the buyer's guide." Multiple CTAs are OK because the visitor's progress is uncertain.

Failure mode

Writing a "best of" list that is transparently a sales pitch for your product. Readers and Google can both spot this. Honest comparison pages, even when they include you, perform far better than disguised brochures.


Transactional intent

The searcher wants to act now. Buy, sign up, download, contact, get a quote.

Signals

  • "Buy," "purchase," "order"
  • "Sign up," "create account"
  • "Download"
  • "Get a quote," "request a demo," "contact sales"
  • Brand + transactional modifier ("buy [product]," "[product] coupon," "[product] discount")
  • "[Product] near me" (local-transactional)
  • "[Product] free trial"

Examples

Query What the searcher wants
"buy iphone 16 pro" Purchase, fast
"salesforce free trial" Trial signup
"download notion mac" Direct download
"request a demo hubspot" Demo form
"plumber near me" A plumber to call
"amazon prime sign up" Amazon Prime signup flow

What to publish

Product pages, pricing pages, signup pages, checkout flows, contact pages. Make the transactional path frictionless. Strong CTAs above the fold. Trust signals near the action.

CTA approach

The CTA is the page. One primary action, sized and placed so the user cannot miss it. Secondary actions allowed if they protect the primary (e.g., "talk to sales" alongside "start free trial" for enterprise vs SMB).

Failure mode

Burying the transactional action behind marketing copy. The user came to act. Get out of the way.


Mixed and ambiguous intent

Many queries do not fit cleanly. Use the SERP to disambiguate.

Method

Open the query in incognito. Look at the top 10 results.

  • All articles? Informational.
  • All branded landing pages? Navigational or transactional.
  • Mix of comparison pages, "best of" lists, and product pages? Commercial investigation.
  • Mix of articles and product pages? Mixed - the safer bet is to produce both kinds of content.

Example: "crm software"

Could be informational ("what is CRM software"), commercial investigation ("which CRM should I pick"), or transactional ("crm software to buy now"). The SERP shows: a definition (HubSpot's), a "best of" list (multiple), several product pages, and a Wikipedia entry. Mixed intent leaning commercial investigation.

The right play: produce a comparison page that opens with a one-paragraph definition. Catches both intents.


Intent shifts across the funnel

The same audience moves through intents.

Stage Example query Intent Right content
Aware of problem "team is missing deadlines" Informational Article on missed-deadline diagnosis
Aware of category "project management software" Informational + commercial Pillar guide
Comparing options "linear vs jira" Commercial investigation Comparison page
Ready to commit "linear pricing" Transactional Pricing page
Already a customer "linear keyboard shortcuts" Informational Help docs

Build coverage across the funnel for your category. A site that only addresses transactional intent has nothing to rank for at the top of funnel and loses the audience to category-defining publishers.


Tagging in research

When pulling keyword research, tag every row.

Query Volume Difficulty Intent Notes
"service blueprint" 8,100 32 Informational Definition + examples piece
"service blueprint vs customer journey" 1,200 28 Commercial investigation Comparison piece
"service blueprint template" 2,400 25 Informational + transactional Template page with download CTA
"service blueprint software" 880 41 Commercial investigation Buyer's guide

The intent tag drives the format and the CTA. Without it, content drifts into whatever the writer felt like producing.


Decision rule

Once intent is tagged:

  • Informational: publish if the topic supports authority-building and the SERP is winnable. Skip if the SERP is dominated by big publishers and you have no asset.
  • Navigational (your brand): make sure you own it. If not, fix.
  • Navigational (someone else's brand): skip.
  • Commercial investigation: publish if you compete in the category. These convert.
  • Transactional: make sure your product / pricing / signup pages rank for the obvious branded transactional queries. Optimize them aggressively.

When intent surprises you

Sometimes a query you assumed was one intent reveals as another. Trust the SERP over your assumption. The SERP reflects what Google has learned about searcher behavior at scale, which is more reliable than your guess.