What striking distance keywords are
Striking distance keywords are search queries where your site already ranks just short of the top results, typically an average position between 8 and 20. Google already considers your page relevant. A focused improvement can be enough to move it onto page one or into the top five.
The term is informal, but the logic is sound: it is far easier to move an existing page from position 12 to position 6 than to rank a brand-new page from nowhere.
Why page two is the cheapest growth
Click-through rate falls steeply with position. Backlinko’s analysis of millions of Google results found the first organic result earns roughly a quarter of all clicks, while fewer than 1% of searchers click anything on page two. AI Overviews and other SERP features have compressed organic CTR further since then, which makes the top few positions even more valuable.
That curve is the opportunity. A query with 5,000 monthly impressions at position 14 might bring in 20 clicks. The same query at position 4 could bring in several hundred. You do not need more impressions; you need a better position for impressions you already have.
Two lists: striking distance and low CTR
There are two different opportunity types in Search Console, and they need different fixes.
| List | Definition | Usual problem | Usual fix |
|---|---|---|---|
| Striking distance | Average position 8–20, meaningful impressions | Page is relevant but not the best answer | Content depth, intent match, internal links |
| Low CTR | High impressions, CTR under about 3% | Snippet does not earn the click | Title, meta description, rich results |
GSC MCP Server’s get_search_opportunities tool returns both lists. Use striking_distance for positions 8–20 and low_ctr for high-impression queries with CTR under 3%.
Finding them in Search Console
The manual way
- Open Performance → Search results and set the date range to the last 3 months.
- Enable Average position and Average CTR.
- On the Queries tab, add a filter for position greater than 7.9 and sort by impressions.
- Export, filter out positions above 20 in a spreadsheet, then look up each query’s ranking page one by one on the Pages tab.
It works, but the export stops at 1,000 rows and the query-to-page mapping is tedious. Larger sites lose most of the long tail.
The AI way
Through the API, GSC MCP Server can read well beyond the interface’s 1,000 rows and pair each query with its page. Ask:
Using the last 90 days on example.com, find striking distance queries (positions 8–20) with at least 200 impressions. For each one, show the ranking page, clicks, impressions, CTR and position. Group the results by page and sort pages by total impressions.
Grouping by page is the important step. Ten striking distance queries pointing at the same article usually mean one rewrite, not ten.
Prioritising with a simple score
Not every striking distance query deserves work. A quick score keeps you honest:
- Potential clicks: impressions × the CTR you would expect at position 3–5, minus current clicks.
- Business value: 1 to 3, where 3 means the query signals buying intent or matches a core product page.
- Effort: 1 to 3, where 1 is a title tweak and 3 is a major rewrite.
Priority = potential clicks × business value ÷ effort. Ask your assistant to apply it:
For those pages, estimate extra monthly clicks if each query reached position 4, assuming a 7% CTR. Then rank the pages by that estimate. Mark pages that target product or pricing intent as high value.
The fix playbook, page by page
Once you have a shortlist, work through each page in this order. Stop when the page clearly answers the query better than what currently ranks above it.
1. Check intent first
Search the query yourself. If the top results are comparison lists and your page is a product page, more words will not help. You may need a different format or a different page. Ask the assistant for all queries the page ranks for; a mix of informational and commercial queries often signals an intent mismatch.
2. Close content gaps
- Use the exact query language in a subheading if it reads naturally.
- Answer the obvious follow-up questions the top results cover and you do not.
- Add specifics: numbers, steps, examples, screenshots, updated dates.
- Remove padding that pushes the answer below the fold.
3. Strengthen internal links
Link to the page from your stronger pages with descriptive anchor text. Use get_search_pages to find your highest-click pages on related topics; they are your best link sources.
4. Make sure you are not competing with yourself
If two of your pages split impressions for the same query, both may stall on page two. Run a keyword cannibalization check before rewriting.
5. Confirm the page is healthy
Run inspect_url to confirm the page is indexed, that Google chose your intended canonical, and that it was crawled recently. A page Google cannot crawl will not benefit from new content.
Fixing high-impression, low-CTR queries
For queries already ranking well with weak CTR, the snippet is usually the problem. Ask:
Show low-CTR opportunities for the last 28 days. For the ten queries with the most impressions, show the current page and suggest a new title under 60 characters and a meta description under 155 characters that match the query intent.
- Lead the title with the query’s main term, then the benefit.
- Match the searcher’s stage: “pricing”, “vs”, “template”, “how to”.
- Check for branded queries. Low CTR on a competitor’s brand name is normal and not worth chasing.
- Segment by device. A title that truncates on mobile can drag CTR down even when desktop looks fine.
Remember that impressions from AI Overviews count in the Web search totals. A query with a large AI Overview may have lower CTR across every result, so compare against your own history rather than generic benchmarks.
Measuring the result
- Log the date and the change for each page.
- Wait at least 28 days after Google recrawls the page.
- Compare the same length of period before and after, per page and per query.
Compare the 28 days since 1 September with the 28 days before for these five URLs. Show clicks, impressions, CTR and position for each page and for its top five queries.
If you report to clients, these before-and-after tables slot straight into the monthly reporting workflow.
Common mistakes
- Chasing single queries. Optimise the page for the cluster of queries it ranks for, not one keyword.
- Using too short a window. Seven days of data is noisy. Use 28 to 90 days.
- Changing too many things at once. If you rewrite the title, the body and the URL together, you will not know what worked.
- Ignoring Bing. Bing data comes from a different audience and feeds ChatGPT Search and Copilot. See why Bing matters for AI search.
Short on time? The 30-minute weekly routine for founders folds a lighter version of this playbook into one weekly session.
Frequently asked questions
What position counts as striking distance?
Most SEOs use average positions from about 8 to 20: the bottom of page one and the whole of page two. GSC MCP Server’s striking_distance strategy uses positions 8 to 20 by default.
How many impressions should a query have to be worth optimising?
There is no universal threshold. Set it relative to your site: a common starting point is the top 20% of striking distance queries by impressions, or a fixed floor such as 100 impressions a month for small sites and 1,000 for large ones.
How long until changes show up in Search Console?
Recrawling can take days to weeks. Compare equal periods of at least 28 days before and after the change, and allow for the two to three day reporting delay.
Is average position in Search Console accurate?
It is the average of your highest position each time you appeared, weighted by impressions. It is accurate for what it measures, but it blends devices, countries and SERP features, so segment by device or country when a number looks odd.
Should I build new pages for striking distance queries?
Usually not. If an existing page already ranks on page two, improving that page is faster than starting a new one. Create a new page only when the query’s intent is genuinely different from the ranking page’s purpose.