How to use AI to find striking-distance keyword wins

This guide is part of our series on AI content strategy.

Striking-distance keywords are queries where your site already ranks in positions 8 to 20, on the bottom of page one or on page two. Google already considers your page relevant. A focused update often moves these terms onto page one, and that usually takes weeks, while a new page can take months. The hard part is choosing which of the hundreds of candidates to work on, and that is where AI helps.

What you need

• Google Search Console, with query and page data for the last 3 months
• Semrush, for search volume and keyword difficulty
• An AI assistant that reads CSV files, such as Claude
• Optional: GA4 key events by landing page, to weight pages that bring leads or sales

Step 1: Export queries with their pages

You need query and page together, so you know which URL is ranking. The Search Console interface shows one or the other at a time and caps exports at 1,000 rows, so use the Search Console API, a Looker Studio report with query and landing page dimensions, or Screaming Frog’s Search Console integration.

Export the last 3 months with: query, page, clicks, impressions, CTR, average position. Then filter to average position 8 to 20 and at least 50 impressions over the period, so you are not chasing queries nobody searches.

Step 2: Add volume and difficulty

Paste the filtered queries into Semrush Keyword Overview in bulk (up to 100 at a time) or use the Keyword Manager, and export volume, keyword difficulty and intent. Search Console impressions tell you what you have seen; Semrush volume tells you the size of the prize if you reach the top three.

Step 3: Have the AI group and score by page

Updates happen at the page level, so the scoring should too.

Attached are (1) Search Console queries ranking in average position 8 to 20 with
their ranking page, and (2) Semrush volume, difficulty and intent for those
queries. Optional (3): GA4 key events by landing page.
1. Join the files on query. Group all queries by ranking page.
2. For each page, calculate: number of striking-distance queries, total Semrush
   volume, total impressions, average position, average difficulty.
3. Estimate extra monthly clicks if each query moved to position 3, using
   these CTR assumptions: position 3 = 10%, current CTR from Search Console.
   Show the formula.
4. Score each page 1 to 5 on: estimated extra clicks, difficulty (lower is
   better), and key events if provided. Show reasoning per page.
5. Flag any query where two or more of our pages rank in the 8 to 20 range.
Sort pages by score. Output CSV.

The CTR figure in the prompt is a rough planning number, not a forecast. Click-through rates vary a lot depending on ads, AI Overviews and other features above the results. What the estimate gives you is a fair way to compare pages against each other.

Do not skip the flag in point 5. If two of your pages share a query, updating either one may not help until you decide which page should own it. Our guide to keyword cannibalization covers that decision.

Step 4: Diagnose the top 10 pages

For each of the top pages, give the AI the page text and the text of the top three results for its biggest query:

Our page ranks #11 for "[query]" and in positions 8 to 20 for these related
queries: [list]. Attached are our page and the 3 pages ranking top 3.
1. Which of our striking-distance queries does our page not mention or answer?
2. What subtopics do all 3 competitors cover that we do not?
3. Is our page's format the same type as theirs (guide, list, product page,
   tool)? If not, say so.
4. Is our title tag aligned with the main query?
5. List the 5 most useful changes in order, each with the section where it goes.
Do not suggest adding keyword repetitions; suggest content that answers
something.

The last line guards against the most common bad advice for striking-distance work, which is to stuff the query into the page a few more times.

Step 5: Make the updates

Typical updates that move striking-distance terms:

• A new section that answers a query the page ranks for but never addresses directly
• An H2 that uses the wording people search with
• An updated title tag that leads with the main query (see our guide to title tag and meta description rewrites, /title-meta-rewrites/)
• Two or three new internal links from related pages, with descriptive anchor text
• Updated facts, examples, screenshots or prices, with the date changed only if the content changed

After publishing, request indexing for the page in Search Console’s URL Inspection tool.

Step 6: Measure after 4 to 6 weeks

Re-export the same queries and ask the AI to compare position, clicks and impressions before and after for each updated page. Keep a log of what you changed. Pages that did not move after two updates may have a deeper problem, such as the wrong page type or stronger competitors, and belong on your content refresh list instead.

Run the striking-distance export monthly. New queries enter the 8 to 20 range all the time, and a short monthly update cycle usually brings gains month after month.