Enter keywords under "Discover search volume and forecasts" in Google Keyword Planner, a free tool in Google Ads, to see average monthly searches. The figure is an approximate 12-month average that groups close variants; very low-volume terms show a dash. Add Google Trends for growth and autocomplete for real phrasing, and decide what to write not just by volume but by whether you can beat the current top 10.
Ways to Check Search Volume and What Each Tells You
Among Google's free tools, only Keyword Planner gives the number of searches itself. Trends, autocomplete, and Search Console show growth, phrasing, or how your own page appears, so combine them.
| Method | What it shows | Watch out |
|---|---|---|
| Google Keyword Planner | Average monthly searches (12-month average, close variants grouped, approximate), 3-month and YoY change, ad competition | Very low-volume terms show a dash. Some accounts see ranges (e.g., 100–1K) |
| Google Trends | Growth, seasonality, and rising related queries | Relative 0–100 values, not counts. Low-volume terms show no data |
| Google autocomplete | Combinations people actually type (words that reveal the problem) | No counts. Only about the top 10 are shown |
| Search Console | How often your page appeared for the query, and its position | Only where your page appeared. At low positions, a small fraction of searches |
| Third-party keyword tools | Bulk autocomplete plus volumes in one view | Each tool's volume source differs. Check which data it uses |
Steps in Keyword Planner
- Sign in to Google Ads and open Tools → Keyword Planner
- Choose "Discover search volume and forecasts" (use "Discover new keywords" to widen related terms)
- Enter one keyword per line or comma-separated, then click Get started
- Check location, language, and date range at the top. The default is the last 12 months
- Read "Avg. monthly searches" together with "Three month change" and "YoY change"
Some accounts see ranges such as "100–1K" instead of exact numbers. Ours has shown ranges at times and exact numbers at others (the figures here were exact, on November 6, 2026). Google Ads Help doesn't state the conditions. With ranges, plan with the lower bound.
Reading the Numbers: Three Pitfalls
1. Close variants are grouped
Google describes average monthly searches as the average number of times a keyword and its close variants were searched. Adding up variants you entered separately can double-count the same searches.
2. It's a 12-month average
By default it's the average of the last 12 months. For example, the Japanese term for "keyword selection" averaged 320 a month, but its 3-month change was −21% and YoY change −88% (Japan, Oct 2025–Sep 2026, checked November 6, 2026). Looking only at the average, you can mistake a declining term for one with demand.
3. A dash isn't zero
Google says keywords with very low volume, or considered sensitive, aren't covered. A dash means "no number available," not necessarily no searches at all. The test below shows an example.
Test: We Checked 20 Keyword Ideas from an AI
On November 6, 2026, we asked the generative AI used to write this article — with no data — for 20 keyword ideas for an article on keyword selection (in Japanese). We checked them in Keyword Planner (Japan, Google, Oct 2025–Sep 2026) and in Google autocomplete. The autocomplete column shows whether the keyword appeared in suggestions for its parent term (e.g., "keyword selection").
| AI-suggested keyword (Japanese) | Avg. monthly | Autocomplete |
|---|---|---|
| 検索ボリューム 調べ方 | 880 | — |
| キーワード選定 やり方 | 70 | Yes |
| 検索ボリューム 目安 | 50 | Yes |
| キーワード選定 初心者 | 40 | Yes |
| ロングテールキーワード 探し方 | 40 | — |
| キーワード選定 ツール 無料 | 20 | Yes |
| キーワード選定 コツ | 20 | — |
| キーワード選定 ai | 10 | Yes |
| キーワード選定 chatgpt | 10 | — |
| キーワード選定 ブログ | 10 | — |
| キーワード選定 手順 | 10 | — |
| キーワード選定 ロングテール | 10 | — |
| キーワード選定 プロンプト | — | Yes |
| キーワード選定 失敗 | — | — |
| キーワード選定 検索意図 | — | — |
| キーワード選定 競合 | — | — |
| キーワード選定 優先順位 | — | — |
| キーワード選定 エクセル | — | — |
| キーワード 難易度 調べ方 | — | — |
| ビッグキーワード ミドルキーワード 違い | — | — |
- 8 of 20 (40%) showed a dash in Keyword Planner. Of the 12 with numbers, only one had 100+ monthly searches
- The biggest one (880/month, "how to check search volume") wasn't a "keyword selection + X" variant. The AI mostly added words to the parent term, but the real demand sat outside it
- "Keyword selection prompt" showed a dash yet appeared in autocomplete: searched, just not often enough to get a number
- Conversely, the 880/month term didn't appear in the top autocomplete for "search volume." Autocomplete shows only about 10 items, so absence doesn't mean no demand
Our call: AI is good at plausible combinations, but whether people search them is another matter. Let AI widen candidates, and don't ask it for volumes. Verify numbers in Keyword Planner and phrasing in autocomplete, and check Google Trends where the two disagree. This check is also why this article's main keyword moved from "keyword selection" to "how to check search volume."
Search Volume Guidelines for Choosing What to Write
| Volume | What to do |
|---|---|
| 100+ per month | A candidate for an article's main keyword. If big sites or tools fill the top 10, shift to a more specific term answered by steps or judgment |
| 10–100 per month | Worth an article if the reader's problem is specific. Cover terms that share the same answer on one page |
| Dash, but in autocomplete | Possibly a new phrasing, or not yet frequent enough to count. Check growth in Google Trends before making it a heading |
| Dash and not in autocomplete | Don't give it its own article. If the content matters, cover it as a heading or FAQ in a related article |
Before the numbers, look at the top 10 results. See what ranks and what's missing (first-hand information, real test results, decision criteria), and write only when you can do better. After publishing, track impressions and position for the term in Search Console. Keyword Planner and Search Console differ because one counts searches and the other counts appearances of your page.
For grouping collected terms and deciding which share one page, see how to use suggest keywords; for reading the purpose behind a query, see the search intent guide.
Common Mistakes
- Using search volumes an AI gave you as-is. You can't confirm the source or period. Use AI to widen candidates and verify numbers in Keyword Planner
- Reading a dash as zero searches. Very low-volume or sensitive terms are simply excluded, and some still appear in autocomplete
- Looking only at the 12-month average. A high average can hide a recent decline. Check the 3-month and year-over-year change too
- Treating YoY changes under ranges (e.g., ±900%) as decisive. They can be an artifact of crossing a range boundary
- Writing in order of volume. Bigger terms have top results filled by big sites. Check the top 10 first and whether you can do better
Related Tools
Sources (checked November 6, 2026)
- Google Ads Help — Use Keyword Planner (very low-volume and sensitive keywords aren't covered)
- Google Ads Help — About Keyword Planner forecasts (average monthly searches include close variants, 12-month default, approximate)
- Google Trends Help — FAQ about Google Trends data (sampled, normalized to 0–100)
FAQ
Can I check search volume for free?▾
Why does Keyword Planner show ranges like "100–1K"?▾
Does a dash (—) mean nobody searches that keyword?▾
How much search volume makes a keyword worth an article?▾
Can I ask ChatGPT or another AI for search volume?▾
Why don't Keyword Planner and Search Console numbers match?▾
Check Autocomplete and Search Volume Together
Enter one term to collect Google autocomplete in bulk, shown alongside Google Ads search volume data (monthly average, trend, CPC, competition). Volumes are available on the Basic plan and above.
See the Search Volume Tool
Written by
今井政和SEO Director / Frontend Developer
SEO Director with 20+ years of web industry experience. Creator of Direbase and the official WordPress plugin "ORECTIC SEO CHECK." Author of a book on web strategy inspired by Edo-era merchant principles.
@imai_directorRelated Articles
You Can't Filter Out Every GA4 Bot | Keeping Direct and Foreign Traffic Spikes Out of Decisions
On our site, over 40% of 28 days of GA4 sessions were bot clusters. GA4 removes only known bots, and data filters can't exclude by country. From a decision we got wrong: three layers of exclusion and how to pick a source-of-truth number for each decision.
Re-picking the Articles in llms.txt Every Month from GA4 AI Referrals | A Four-Month Log
Four months of rebuilding the "most-referenced by AI" list in llms.txt from GA4 AI referrals. Over half the list turned over each month, and picking by count alone let in off-topic pages. What GA4 AI referrals do and don't count, and why we can't claim an effect.
SEO Strategy for Tool-Based Content | Design Patterns That Match Search Intent
Why do tool-based pages outperform articles in search? Using real CTR and engagement data, we demonstrate the cross-tool analysis process with GA4, Direbase, and GSC.
How to Research Co-occurring Words: Compare Top Pages and Find Missing Topics
Compare top pages with your own in a co-occurrence tool, strike theme words, filler, and frame noise, and find only the topics you're missing — with a measured comparison of five top pages.