Explore weighted concepts around the search phrase Query Expansion examines related concepts that the Similar Words system associates with the entered search phrase. Each concept carries a relationship strength, such as 0.7 or 0.9. That number describes how strongly the expansion is connected to the source query in this configured set; it is not a keyword-density goal or a demand that every term appear. Load Similar Words to retrieve existing relationships or lazily generate missing suggestions through the established fallback. Review the resulting list before interpreting a score. Remove concepts that represent the wrong meaning, audience, industry, or intent. A search phrase can be ambiguous, and an expansion that is statistically related may be inappropriate for the target page. The editor remains responsible for deciding which relationships belong in an accurate answer. Better inputs produce a more useful heatmap than an indiscriminate collection of synonyms.
Read color intensity as relationship strength The heatmap marks only complete related-concept phrases that contribute to the calculation. All words in a multiword concept must occur together, in order, within one scored field; words scattered between the title and body do not create a match. Darker color corresponds to a stronger configured relationship weight, so a 0.9 match is more prominent than a 0.5 match. Hover a mark to see the complete concept and its weight. Each field displays its own Related-concept coverage score: the relationship weight found in that field divided by all eligible relationship weight. The page-level Query Expansion Score de-duplicates those matches. If the same concept occurs in both the title and visible content, both field scores show it, but its weight enters the page numerator only once. The displayed formula exposes the matched numerator, total denominator, and resulting percentage.
Turn an absent concept into an answer Do not paste the related-concepts list into the page. Group missing concepts into meaningful themes: questions a visitor asks, attributes they compare, prerequisites they need, examples that clarify the topic, or decisions they must make next. Choose one theme that fits the page and write a coherent passage. The concept should appear because the new information requires it, not because a colored meter was incomplete. Some expansions are interchangeable vocabulary rather than missing subjects. If the current wording is clearer, leave it alone. Other expansions expose a genuine information gap. For a program page, related ideas might identify training format, certification, coursework, or career outcomes. Confirm every claim against authoritative business information. Query expansion can suggest what to investigate, but it cannot determine what the organization actually offers.
Use showdown mode to compare concept breadth A comparison URL displays the same weighted concept set against both pages. Inspect the paired field scores, heatmaps, and Related Concept Scores Showdown to see which complete phrases are present and exactly which fields supply them. The outperformer may cover a useful subtopic your target omits, while your target may contain stronger evidence for another concept. The comparison is most useful as a map of differing subject coverage. Keep the weights constant while comparing. Changing the expansion list midway creates a different measurement and makes the bars incomparable. Do not assume that every concept unique to the outperformer belongs on your page; the documents may serve different intents. Select only a gap that supports the target’s purpose, then express it with original language, evidence, and an appropriate level of detail.
Measure how a draft changes weighted coverage Edit the title, description, path preview, or visible copy and click Measure This Change. The edited field score, page formula, concept cards, and heatmap update from the supplied fields. A concept contributes once until its last complete occurrence disappears from every scored field. Removing sentences that contain no last occurrence correctly leaves the page score unchanged, and the result message identifies that reason. To add a match, place the complete concept naturally within one field. A successful experiment should make the page more complete, not merely more colorful. Read the new section without looking at the heatmap and ask whether it answers a real question. Remove redundant synonyms and unsupported tangents. Save the relationship set used for the test so later measurements remain comparable. Once the revision is published and recrawled, verify that the stored page retains the intended concept evidence.
Maintain a trustworthy expansion set Related language changes with products, audiences, and search behavior. Review saved concepts periodically rather than treating one generated list as permanent truth. Remove obsolete brand terms, ambiguous abbreviations, and relationships that lead away from the page’s purpose. Add a missing concept only when there is evidence that searchers use it or readers need the information it represents. Keep the strongest, clearest relationships at meaningful weights so the heatmap remains legible and the score responds to editorially important coverage. Weight changes alter the definition of the factor. If “certification requirements” moves from 0.5 to 0.9, a page containing that concept receives more credit even though its content did not change. Document adjustments and avoid comparing scores from different weight sets without noting the revision. During a page comparison, use one frozen set for both URLs. This makes a showdown a fair view of content evidence rather than a mixture of content differences and configuration changes. Expansion analysis also supports content planning beyond a single edit. Cluster accepted concepts into sections that belong on the target, separate subjects that deserve supporting pages, and connect those pages with descriptive internal links. Reject concepts that would duplicate another page’s intent. This creates a clearer information architecture while preserving topical focus. The explorer is doing useful work when it helps decide where an idea belongs—even when the decision is to cover that idea elsewhere and leave the target’s Query Expansion Score unchanged.
