How to Find Striking-Distance Keywords With AI in 45 Minutes
Your next meaningful SEO win may already be hiding on your website.
Pages ranking just outside the strongest organic positions often need less work than brand-new content. Backlinko’s analysis of four million Google results found that the first organic result earns an average click-through rate of 27.6%. It also found that moving up one position increases relative CTR by 2.8% on average, although the gains vary substantially across the results page (Backlinko).
Striking-distance keyword analysis helps you find pages that may be close to earning more visibility and clicks. AI can accelerate the process by organizing exports, identifying patterns, comparing intent, and suggesting improvements. It cannot decide which opportunities genuinely deserve your time, so the final judgment remains yours.
What Are Striking-Distance Keywords?
Striking-distance keywords are search queries for which your site already ranks reasonably well but has not yet reached its full traffic potential.
SEO teams commonly define them as keywords ranking between positions 4 and 20. A tighter analysis might focus on positions 6 to 15, while a newer or less authoritative site may treat positions 11 to 30 as realistic opportunities.
There is no universal range. A useful definition depends on:
- Your site’s current authority
- The competitiveness of your market
- Search volume and business value
- The page’s existing performance
- The search features occupying the results page
- How closely the content matches current intent
A query in position 12 with thousands of impressions and a highly relevant landing page may be a strong opportunity. A query in position 7 with ten impressions, weak commercial relevance, and an unsuitable page probably is not.
Google Search Console reports position as the average position of your site’s topmost result for a query. Google recommends focusing on trends in impressions and clicks rather than treating position as an isolated measure (Google Search Console Help).
That distinction matters. Average position can change because of location, device, result features, query variations, and the mix of searches in which your page appeared.
How AI Helps With Striking-Distance Keyword Research
A normal Search Console export can contain hundreds or thousands of queries. Reviewing each row manually is slow, and sorting only by position produces a misleading priority list.
AI helps you process the data more systematically. It can:
- Filter queries into realistic ranking ranges
- Group related terms by topic and intent
- connect multiple queries to one underlying content problem
- Flag pages with high impressions but weak CTR
- Detect likely intent mismatches
- Generate hypotheses about missing sections
- Build a prioritized optimization queue
- Summarize opportunities for non-SEO stakeholders
The key word is hypotheses. An AI model does not automatically know why a page ranks in position 11. It may not see the live results page, your analytics, your conversion data, or recent changes to competing pages.
Use AI as an analyst that organizes evidence, not as an oracle that invents certainty.
What You Need Before Starting
You only need a few basic inputs:
- Access to Google Search Console
- A spreadsheet application
- An AI tool that can analyze uploaded tables
- Your target market and audience
- A list of important products, services, or conversions
- Optional access to a keyword or rank-tracking platform
Use at least 90 days of Search Console data when possible. A shorter period may be appropriate for seasonal campaigns, recent launches, or sites that have changed substantially.
Before uploading an export to an external AI service, review its privacy and data-retention terms. Remove confidential URL parameters, customer information, unpublished product names, and any other sensitive data.
The 45-Minute AI Workflow
Minutes 0–5: Export the Right Search Console Data
Open the Performance report in Google Search Console and select:
- Search type: Web
- Date: Last three months
- Metrics: Clicks, impressions, average CTR, and average position
Export the query and page data. If practical, create a combined table containing:
| Query | Landing page | Clicks | Impressions | CTR | Position |
|---|---|---|---|---|---|
| seo content brief template | /content-brief-guide/ | 34 | 2,800 | 1.2% | 8.7 |
| ai keyword clustering | /keyword-clustering/ | 18 | 1,450 | 1.2% | 11.4 |
The landing page is essential. A keyword without its ranking URL tells you that an opportunity exists but not what you need to improve.
Avoid mixing branded and non-branded searches at this stage. Branded queries frequently have different CTR and conversion patterns, which can distort prioritization.
Minutes 5–10: Clean and Filter the Export
Remove irrelevant or unreliable rows, including:
- Queries with extremely low impressions
- Navigational searches for your brand
- Queries in languages or countries you do not serve
- Obvious tracking noise
- Queries already ranking consistently in the top three
- Terms unrelated to the ranking page’s purpose
Set a minimum impression threshold that fits your site. A large website might require 500 impressions per query, while a smaller B2B site may find 50 impressions meaningful.
Next, filter for your preferred position range. A practical starting point is:
- Positions 4–10: CTR and snippet opportunities
- Positions 11–20: page-one breakthrough opportunities
- Positions 21–30: possible opportunities requiring larger changes
Do not combine all three groups into one score. Moving from position 19 to position 8 is usually a different project from improving position 5 to position 3.
Minutes 10–18: Ask AI to Classify the Opportunities
Upload the cleaned file and give the model enough context to avoid generic output.
A useful prompt is:
Act as an SEO analyst. Analyze the attached Google Search Console data.
For each query and landing-page pair:
1. Classify the likely intent as informational, commercial, transactional, or navigational.
2. Group semantically related queries that rank to the same URL.
3. Flag likely intent mismatches between the query and landing page.
4. Identify high-impression queries with below-average CTR.
5. Separate opportunities into positions 4-10, 11-20, and 21-30.
6. Do not invent search volume, ranking factors, or SERP features.
Return a table with:
query cluster, landing page, total impressions, clicks, weighted position,
likely intent, opportunity type, confidence level, and recommended next check.
Explicitly tell the model not to invent missing metrics. Otherwise, it may produce plausible-looking search volumes or competitor observations that are not present in your file.
Ask it to show uncertainty as high, medium, or low confidence. This makes the output easier to review critically.
Minutes 18–25: Build an Opportunity Score
A simple score prevents you from prioritizing keywords based on impressions alone.
Score each cluster from one to five across these factors:
| Factor | Question |
|---|---|
| Visibility | Does the cluster receive meaningful impressions? |
| Proximity | How close is it to the desired ranking range? |
| Relevance | Does the query support a product, service, or business goal? |
| Intent fit | Is the existing page appropriate for the query? |
| Effort | Can you improve the page without rebuilding it? |
You can calculate a basic score with this formula:
Opportunity score =
Visibility + Proximity + Relevance + Intent fit + Ease of improvement
For a more business-focused model, give relevance and intent twice the weight:
Opportunity score =
Visibility + Proximity + (Relevance × 2) + (Intent fit × 2) + Ease
AI can assign an initial score, but you should challenge it. The model does not know that one service has high margins, another is being discontinued, or a low-volume term regularly generates qualified leads.
Minutes 25–33: Verify the Live Search Results
Now inspect the current Google results for your highest-scoring clusters. This is where you test the AI’s assumptions.
Check:
- What content type ranks: guide, category, product, video, tool, or forum
- Whether the results satisfy informational or commercial intent
- How recently competing pages were updated
- Whether an AI Overview, featured snippet, local pack, or video block appears
- Whether your title is less specific than competing titles
- Whether competitors provide information your page lacks
- Whether the ranking URL is the best page on your site for that query
This manual review is increasingly important because search result layouts are changing. Semrush analyzed more than ten million keywords and found that Google AI Overviews appeared for 6.49% of queries in January 2025, peaked at 24.61% in July, and settled at 15.69% in November (Semrush).
The same study found that over 88% of queries triggering AI Overviews had informational intent (Semrush). A traditional blue-link position therefore does not reveal the page’s actual visual prominence or click potential.
Record what you observe instead of asking AI to guess what is on the live results page.
Minutes 33–40: Map Each Keyword Cluster to a Fix
Most striking-distance opportunities fall into a small number of categories.
CTR problem
The page has good visibility but earns fewer clicks than expected.
Possible improvements include:
- A clearer title tag
- Stronger alignment between the title and query
- A more specific value proposition
- Updated dates where freshness genuinely matters
- A more useful meta description
- Better structured data where eligible
Content-depth problem
The page matches the basic intent but omits an important subtopic.
Possible improvements include:
- Answering related questions
- Adding definitions, steps, examples, or comparisons
- Supporting claims with original evidence
- Clarifying sections that currently give shallow answers
- Adding expert commentary or first-hand experience
Intent problem
The page is not the right format for the query.
A product page may struggle for an informational search, while a long guide may be wrong for a transactional query. The solution could be a new page, a substantial repositioning, or a decision not to target that term.
If rankings changed because the results now favor a different intent, use a structured How to Audit Search Intent Drift With AI in 45 Minutes before rewriting the page.
Internal-link problem
A relevant page exists but receives little contextual support from the rest of the site.
Add links from closely related, authoritative pages using natural anchor text. For a repeatable process, see this guide to How to Build AI-Driven Internal Links in 30 Minutes.
Cannibalization problem
Two or more URLs compete for the same query cluster.
Compare their purpose, links, conversions, and ranking history. You may need to consolidate overlapping content, clarify each page’s target, or improve internal linking. Do not merge pages solely because an AI tool labels them similar.
Minutes 40–45: Create a Short Action Queue
Finish with three to five opportunities rather than a list of 100 keywords.
Your queue should include:
| Priority | Page | Keyword cluster | Evidence | Planned improvement |
|---|---|---|---|---|
| 1 | /content-brief-guide/ | SEO content brief templates | Position 8.7, 2,800 impressions, weak CTR | Improve title and add downloadable example |
| 2 | /keyword-clustering/ | AI keyword clustering | Position 11.4, strong intent fit | Expand workflow and add internal links |
| 3 | /seo-audit/ | automated SEO audit | Two URLs ranking | Review cannibalization |
Include a baseline date and the metrics you will compare later. Depending on crawl frequency and competition, evaluate meaningful trends after approximately four to eight weeks rather than expecting an immediate response.
A Prompt for Turning Findings Into Page Recommendations
Once you have verified the results page, give AI your evidence rather than requesting a blind rewrite.
Create an optimization brief for this existing page.
Target query cluster:
[queries]
Current Search Console metrics:
[metrics]
Observed search intent:
[intent]
SERP observations:
[notes]
Current page outline:
[outline]
Recommend:
1. Sections to retain
2. Missing information to add
3. Sections that do not support the target intent
4. Title and heading improvements
5. Relevant internal-link opportunities
6. Claims that require first-party evidence or external sources
Do not rewrite the full article. Do not recommend adding text solely to increase word count.
This produces a controlled editing brief instead of an unnecessarily long AI draft.
Google explicitly says, “Generative AI can be particularly useful when researching a topic, and to add structure to original content” (Google Search Central). The same guidance warns that generating many pages without adding value may violate its scaled content abuse policy.
For AI-assisted updates, verify facts, add genuine experience, and run proper Stop Publishing AI Content Without These SEO Checks.
Practical Tips That Improve the Analysis
Work With Clusters, Not Isolated Keywords
One page often ranks for several variations of the same need. Treating each variation as a separate project creates repetitive content and possible cannibalization.
Cluster terms when they share:
- The same intent
- Similar wording or entities
- The same ranking page
- Substantially similar search results
- One answer that would satisfy the searcher
Separate Position Gains From CTR Gains
A page in position 5 may not need more content. It might need a clearer title or a richer result. Conversely, a page in position 14 usually needs stronger relevance, authority, depth, or internal support.
Diagnose the problem before choosing the fix.
Compare Devices and Countries
A query may rank well on desktop and poorly on mobile, or perform differently between countries. Segment high-priority opportunities in Search Console when your initial averages look inconsistent.
Protect What Already Works
Do not replace a successful page with a completely AI-generated version. Preserve sections that attract links, rank for valuable terms, demonstrate experience, or convert visitors.
Add what is missing and remove only what clearly conflicts with the page’s purpose.
Track Conversions, Not Just Rankings
More clicks are not automatically more valuable. Connect your shortlist to leads, sales, subscriptions, or assisted conversions wherever possible.
A query with 200 impressions and strong buying intent may deserve attention before a broad informational term with 10,000 impressions.
Pros and Cons of Using AI
Advantages
- Faster data processing: AI can classify large exports in minutes.
- Better pattern detection: It can identify clusters and recurring content gaps.
- Consistent prioritization: A fixed scoring framework reduces arbitrary choices.
- Accessible analysis: Beginners can turn raw Search Console rows into understandable categories.
- Reusable workflows: Prompts and scoring models can support monthly reviews.
Limitations
- Possible hallucinations: AI may invent volumes, competitors, or SERP features.
- Weak business context: It cannot infer your margins or strategic priorities without input.
- Imperfect intent classification: Similar queries can require different content formats.
- Privacy concerns: Search and conversion data may be sensitive.
- False precision: A detailed score is still an estimate, not a ranking forecast.
- Over-optimization risk: Applying every suggestion can make content repetitive and less useful.
Human review is not an optional final polish. It is what turns automated pattern matching into an SEO decision.
How Striking-Distance SEO Is Evolving
Striking-distance analysis used to focus heavily on reaching page one. That boundary is less meaningful on results pages containing AI answers, videos, discussions, shopping modules, and other features.
Modern analysis should therefore consider three kinds of visibility:
- Ranking visibility: Your conventional organic position
- SERP visibility: How prominent your result looks on the page
- Answer visibility: Whether your information or brand appears in AI-generated answers
This does not make traditional rankings irrelevant. It means position data needs context. Google itself advises site owners to prioritize helpful, reliable, people-first information rather than content created mainly to manipulate rankings (Google Search Central).
The most durable striking-distance improvements usually make an existing page more complete, specific, trustworthy, and aligned with what the searcher needs.
Final Thoughts
A 45-minute striking-distance keyword review combines Search Console evidence, AI-assisted classification, live SERP verification, and business judgment. The goal is not to generate the longest keyword list. It is to find a few existing pages where a focused improvement has a credible chance of producing better visibility, qualified clicks, or conversions.
AI shortens the analysis. Your evidence, priorities, and editorial decisions determine whether the resulting work is useful.