Co-occurring Words: Find the Topics You're Missing

What Co-occurring Words Are, and What Google Says

Co-occurring words are words that tend to appear in the same text as a given term. In SEO the phrase usually means "words that show up often on the pages ranking for your term." It isn't a term Google uses in its documentation, so start from the premise that it isn't a named ranking factor.

Google does say three relevant things. First, in How Search Works, the most basic relevance signal is content containing the same keywords as the query — and algorithms also check for other relevant content, like pictures of dogs, videos, or a list of breeds for a search about dogs. Second, its spam policies call it keyword stuffing to fill a page with keywords to manipulate rankings, including repeating words so often it sounds unnatural. Third, its helpful-content guidance says Google has no preferred word count, while asking whether content gives a substantial, complete, or comprehensive description of the topic.

How to Research Co-occurring Words: 5 Steps

1. 1. Search your target term and pick about five pages with the same intent

   Go down the results and keep only pages that answer the way your article will. Leave out tool pages, official documentation, product manuals, and store pages. Mixing in pages that answer differently fills the table with their words.

2. 2. Run the chosen URLs and your own page through a comparison

   Always include your own page — listing only the top pages can't tell you what you're missing. Direbase's co-occurrence comparison takes up to 10 URLs at once and shows each word's count per page plus the total.

3. 3. Strike the words that carry no information

   Four kinds: (1) theme words that appear on every page — your target term itself; (2) filler words; (3) words from the page frame — company names, menus, article lists, careers links; (4) words from code or URL samples (www, com, example). The top of the table is usually full of these.

4. 4. Group what's left into topics

   "Video," "image," and "news" become "image and video sitemaps"; "plugin" and "wordpress" become "building one in WordPress." Don't chase single words — read off which sections the top pages have.

5. 5. Check whether your page has that section, then decide

   Judge by whether the section exists, not by counts. If the topic is missing, add an explanation only if your readers need it. If the topic is there but the word just doesn't rank high in the table, do nothing — shorter pages show lower counts for the same topic.

The pages from step 1 work for heading comparison too — see how to build an SEO article outline. For choosing the target term, see how to use suggest keywords.

Measured: Five Top Pages for an XML Sitemap Query vs. Our Article

On October 30, 2026, we searched Google (Japanese) for "xml sitemap" and picked five explainer articles (A–E), leaving out a generator site, a product manual, and Google's own documentation. We ran those five plus this site's XML sitemap article (around position 90 for that query) through Direbase's co-occurrence comparison (Pro plan; each page's top 50 words).

The five pages yielded 143 distinct words. The 9 found on all five were theme words plus one filler word — no information there. The shortest page, D (about 2,900 characters including the frame), had 28 of its top 50 words from company names, article lists, careers links, and filler — more frame than topic. The longest, A (about 31,000 characters), had www, com, and example near the top, all from code samples. A co-occurrence table reflects page length and format, not just topic coverage.

Of the 30 words found on three or more pages, 11 weren't in our article's top 50. Only four of those name a topic ("create," "register," "tool," "role"), and a count of the body showed "create," "register," and "tool" five times each. The sections on building a sitemap, submitting it in Search Console, and tools were already there; the article is short (about 6,100 characters), so the counts didn't reach the top 50. Nothing to add here.

The candidates were "video" and "news," found on two pages (A and C), plus "image" on A alone — the topic of image, video, and news sitemaps, which our article doesn't cover. But our readers are web practitioners, and most sites don't need those extensions, so we noted it as a short "which sites need this" addition for the next revision. Meanwhile, the words only our article had — lastmod, priority, changefreq, Search Console — reflect its angle: tag-by-tag calls on how far to trust lastmod and what to do with priority and changefreq. We won't trim that to look like the top pages.

Decide by How a Word Is Spread

Add a chosen topic as a section in the existing article first; a new article can end up competing with your own page for the same search. For rewriting, see how to rewrite an article; for reading intent, see how to identify search intent.

Common Mistakes with Co-occurring Words

• Adding words to match the top pages' counts. No new topics — just unnatural repetition
• Feeding in the top 10 as-is. Tools, official docs, and manuals mix in, and the comparison stops being article to article
• Comparing without your own page. You learn what the top pages share, not what you lack
• Reading a long page's words as "important." High counts can just reflect length and code samples
• Cutting words only you have. That difference is the reason to read your article

Related Tools

• Keyword Research — co-occurrence comparison, headings, suggestions
• Competitor Keywords — see what competitors rank for
• SEO Score Check — titles, headings, and the basics

Sources (checked October 30, 2026)

• Google — How Search Works: Ranking results (relevance of content)
• Google Search Central — Spam policies (keyword stuffing)
• Google Search Central — Creating helpful, reliable, people-first content
• Google Search Central — Learn about sitemaps (official reference for the test topic)