Commercial Intent Keywords: How to Classify and Prioritize by Buying Stage
- Filed November 10, 2026
- Keyword research, guide 2 of 6
A commercial intent keyword is a search made by someone moving toward a purchase: comparing options, checking prices, looking for a provider. Classifying intent correctly decides what kind of page should answer the query, and getting it wrong is costly in both directions: a sales page for a research query does not give the searcher what they came for, and an educational article for a ready-to-buy query wastes the visit. The trap is classifying by a single word. “Best”, “price” or “near me” suggest intent, but the results page and the full query decide it.
Intent types by buying stage
| Stage | Intent | Example queries | Page type that fits |
|---|---|---|---|
| Problem awareness | Informational | why is my water heater making noise | Explanatory article |
| Solution research | Informational to commercial | tankless vs tank water heater | Comparison guide |
| Provider or product comparison | Commercial investigation | best tankless water heater brands, rinnai vs navien | Comparison, reviews, buyer's guide |
| Service evaluation | Commercial (local) | water heater installation cost, plumber reviews | Pricing explainer, service page with costs |
| Purchase or hire | Transactional | buy rinnai re199, emergency plumber near me | Product page, service page with booking |
| Post-purchase | Navigational or support | rinnai error code 11 | Support article, manual |
Why one modifier is not enough
| Query | Modifier | Likely intent (confirm on the results page) |
|---|---|---|
| best way to clean a coffee grinder | "best" | Informational how-to |
| best coffee grinder under 100 | "best" | Commercial investigation |
| coffee grinder price | "price" | Commercial; check whether shopping listings appear |
| price of coffee per pound history | "price" | Informational |
| plumber near me | "near me" | Transactional, local |
| plumber near me salary | "near me" | Informational, about jobs |
The modifier narrows the possibilities; the rest of the query and the results Google shows settle them.
An intent classification decision matrix
Classify each keyword in four steps.
1. Read the query as a whole. What task would a person typing these exact words be trying to complete?
2. Look at the results page. Search the query (signed out, in the target location) and note:
- the page types ranking (product pages, category pages, comparisons, guides, local results, videos);
- features present (shopping listings, local pack, AI Overview, “People also ask”);
- whether results mix types, which signals mixed intent.
3. Assign a primary intent and stage, using the table above.
4. Score commercial value, combining the stage with what the business sells:
| High relevance to what you sell | Low relevance | |
|---|---|---|
| Transactional or commercial investigation | Priority 1: dedicated product, service or comparison page | Priority 3: only if it supports a broader audience goal |
| Informational | Priority 2: guide that links to commercial pages where useful | Priority 4: skip unless it serves a broader audience goal |
Mixed intent
Some queries serve several intents at once. “Standing desk” might show product listings, a comparison article and a buying guide together. Options:
- Build the page type that dominates the results, and cover the secondary intent within it (a category page with a short buying guide section).
- Or build two pages for two distinct queries (“standing desk” for the category, “how to choose a standing desk” for the guide), making sure each owns a different query and links to the other.
Avoid building two pages for the same query and intent; they compete with each other.
Prioritizing a keyword list
For each keyword, record: query, stage, intent, page type indicated by the results, current ranking page (if any), commercial relevance, an estimated demand figure with its source and, if you use one, a difficulty score from a single tool. Then prioritize:
- Transactional and commercial investigation queries highly relevant to what you sell, where you have no suitable page.
- The same queries where you have a page of the wrong type.
- Informational queries that lead naturally to your products, where a guide can serve the reader and point to commercial pages.
Demand estimates come from keyword tools and are uncertain; use them to compare, not to forecast exact traffic.
A worked classification example
As an illustration, take a company that sells and installs solar panels in Arizona, reviewing five keywords.
| Keyword | Results page shows | Intent and stage | Decision |
|---|---|---|---|
| solar panel installation phoenix | Local pack, installer sites | Transactional, local | Priority 1: service page for Phoenix with real service details |
| are solar panels worth it in arizona | Guides, news, an AI Overview | Informational to commercial | Priority 2: guide with local cost and incentive facts, linking to the service page |
| best solar panels 2026 | Review and comparison sites | Commercial investigation | Priority 1 by the matrix, but only if the company can write an honest comparison including brands it does not install; otherwise skip |
| solar panel cleaning tips | How-to articles, videos | Informational, post-purchase | Priority 4 unless the company offers cleaning |
| tesla solar roof cost | Brand pages and cost articles | Commercial, brand-specific | Skip unless the company installs that product |
The labels in the table are illustrative; confirm each by checking the actual results page in the target location.
Classification mistakes
- Labeling by modifier alone.
- Treating every query with commercial value as transactional.
- Ignoring local results that show the query is location-dependent.
- Assigning one page to several intents and serving none well.
- Writing disguised sales pages for informational queries, which leaves the reader’s actual question unanswered.
Intent classification is a reading exercise: the words, the results and the business’s offer together tell you what page should exist. A matrix makes those readings consistent across a large keyword list.