Skip to content

Stuck on Page Two of Google? The Prompt That Diagnoses What’s Holding Each Page Back (Full Prompt Included)

Add us as a preferred source on Google

Stuck on page two of Google? Filter Search Console to positions 8-20, upload to Claude, and see what's likely blocking each page. Free full prompt included.

Omri Kfir, DigiAssets team Omri Kfir · Updated · 11 min read
Stuck on Page Two of Google? The Prompt That Diagnoses What’s Holding Each Page Back (Full Prompt Included)

If your site is stuck on page two of Google, it’s already showing you why. In Search Console, filter average position between 8 and 20, sort by impressions, and you get the pages closest to page one. One prompt in Claude diagnoses each page and tells you what to fix first. The full prompt is below.

A page stuck on page two of Google is a page Google already knows and ranks for relevant search terms, but at an average position of roughly 8 to 20, which puts it below the organic results on page one. The foundation is already there, so it’s usually the easiest and fastest content to improve on your site.

“If you’re on page two, Google knows you very well. It has already decided you’re relevant. It just isn’t convinced enough to move you up.”

Omri Kfir, Co-founder, DigiAssets

What prompt diagnoses why a page is stuck on page two?

This is the prompt we use at DigiAssets to diagnose pages stuck on page two. Export the pages from Search Console, open Claude, upload the file and paste the prompt. Fill in the fields in brackets, and if you have the URLs of the top three competitors for each important keyword, add those too.

The prompt:

Prompt to copy

You are a senior SEO strategist. I’m uploading a CSV file from Google Search Console with pages and search queries where my site ranks at an average position between 8 and 20.

My input:
1. My site: [domain] and niche: [short description of the business and audience]
2. Attached file: [upload a CSV of Queries and Pages here, with impressions, clicks, CTR and position]
3. Optional: URLs of the top 3 competitors on Google for each important keyword: [paste here, or write “none”]

Step 1, filtering and prioritizing:
Group the keywords by the page that ranks for them. Rank each page by potential: high impressions, position close to 10, and low CTR relative to position. Show the 15 highest-priority pages in a table.

Step 2, diagnosis for each page:
For each page in the table, check the following four possible causes and rate each one as likely, possible or unlikely:
A. Search intent mismatch (the page is a different type of content than the keyword calls for)
B. Coverage gaps (questions, subtopics or proof missing from the page)
C. Weak internal linking (the page doesn’t get enough links from relevant pages)
D. Title and description problem (a title or meta description that doesn’t earn the click)

If I provided competitor URLs, compare my page to them and infer coverage and intent from them. If I didn’t, write at the top of your answer: “This diagnosis is based on the keywords and data only, with no competitor comparison, so it is less reliable.”

Step 3, action:
For each page, give:
1. The most likely reason it’s stuck, and why
2. The smallest change that could move it
3. A second change if the first isn’t enough
4. What to check on the page itself to confirm the hypothesis

Step 4, work order:
Build a two-week plan. Order the pages from the cheapest, fastest change to the most expensive one. For each task, give an effort estimate: low, medium or high.

Rules:

  • Don’t make up data that isn’t in the file.
  • Don’t claim you’ve seen the content of my page or a competitor’s page unless I pasted it. If you’re inferring something from a URL or a search query alone, write “inferred from URL only”.
  • Clearly separate: observed in the data, your estimate, and assumption.
  • Every diagnosis is a hypothesis to test, not a verdict. Don’t promise ranking gains or a timeline for results.
  • Don’t suggest changing titles or URLs of pages that already rank well without warning about the risk.
  • If the problem looks like overall site authority rather than a page-level issue, say so explicitly.
  • Write in English, in direct language and short sentences. No em dashes.
  • End with a single table: page, main keyword, likely cause, first change, effort, priority.

This prompt is also available in the DigiAssets prompt library, with a one-click copy button and more prompts for SEO and GEO.

Want to know which pages on your site are closest to page one? Get a free website review →

Why is a page on page two the easiest opportunity on your site?

Because the foundation is already there. Most people think the problem is that Google doesn’t know them. But if you’re on page two, Google knows you very well, and it has already decided you’re relevant. It just isn’t convinced enough to move you up.

A new page starts from zero: it has to be crawled, get indexed, and prove to Google that it’s relevant to the keyword at all. A page already sitting at position 11 or 14 has cleared all of that. If it were that bad, it wouldn’t even be on page two. That’s why it’s the content that will be easiest and fastest for you to improve.

The hard part isn’t finding the list. It’s understanding why each page got stuck under this glass ceiling. Checking every possible cause by hand, page after page, can take hours. That’s why we built the prompt.

How do you find the pages stuck on page two in Search Console?

Go to Google Search Console, open the Performance report, filter average position between 8 and 20, and sort by impressions. What you get is a list of the pages and keywords where you’re already very close to page one.

  1. Open the Performance report. In the menu, choose Performance, then Search results. Pick a 3-month range so the average isn’t driven by a single week.
  2. Turn on the Average position metric. Click its card above the chart so the column shows up in the table. Google defines it as the average position of the topmost result from your site (Search Console Help).
  3. Filter position between 8 and 20. The table lets you filter rows by metric, for example clicks below a certain number (Search Console Help). Filter Position greater than 8, and if the interface doesn’t allow a second condition on the same metric, filter out the upper bound (20) after exporting.
  4. Sort by impressions. Click the Impressions column header. The keywords and pages with the most potential will be at the top.
  5. Export to a file. The export button downloads the chart and table data. Upload the file to Claude along with the prompt.
Pages on Google's page two stuck under a glass ceiling and diagnosed with Search Console data
Pages on page two are already relevant. The question is what’s holding them back

Pro tip: Check position in the Pages tab, not just the chart. According to Google, the chart calculates position from the site’s single highest result, while the table calculates it per URL. So per-page position and impressions can differ from the overall number (Search Console Help).

What are the four reasons a page gets stuck on page two?

Sometimes the page answers a different search intent. Sometimes it lacks internal links. Sometimes the competitors on page one cover questions your page doesn’t answer. And sometimes the title just doesn’t earn the click. The prompt checks all four for every page and rates each as likely, possible or unlikely.

Cause What it looks like What to check to confirm
Search intent mismatch An informational article ranking for a keyword where people are looking for a provider, or the other way around What type of pages are on Google’s first page for that keyword
Coverage gaps Questions, subtopics or proof that competitors address and you don’t Comparison with the top three competitors
Weak internal linking The page gets almost no links from relevant pages on the site How many pages on the site link to it, and with what anchor text
Title and description problem Lots of impressions, low CTR relative to position How the title and description look in the results next to competitors

The last two causes are backed directly by Google’s documentation. On links, Google says it uses links to determine the relevance of pages, and that every page you care about should have a link from at least one other page on your site (Google Search Central). On titles, Google recommends a unique, descriptive and concise title for every page (Google Search Central).

How do you work with Claude’s diagnosis?

The prompt takes each page and tells you what is most likely holding it back, what to fix first, and the smallest change that could move it up. At the end it builds a two-week plan, ordered from the cheapest, fastest change to the most expensive, with an effort estimate for each task.

  • Start with the cheap fixes. An internal link from a relevant page or a sharper title takes minutes. Rewriting a page takes days. If the cheap change moves the page, you’ve saved yourself the expensive one.
  • Every diagnosis is a hypothesis. The prompt requires Claude to state what you need to check on the page itself to confirm it. Check before you change anything.
  • Be careful with pages that already rank well. Changing the title or URL of a page that drives traffic on other keywords can cost you those rankings. The prompt is required to warn you about this.
  • Want a more accurate diagnosis? Give it competitors. Without competitor URLs, it will state at the top of its answer that the diagnosis is less reliable. To see what competitors do on the pages that bring them leads, we also have a prompt that breaks down your competitors’ money pages.

When we at DigiAssets take on a new site, pages on page two are among the first things we check, because that’s where the smallest change can have the fastest impact. And if you want to understand what Google looks for in the page that ranks first, we wrote about it in Do you really deserve to rank first on Google?.

When won’t the page-two diagnosis work?

The method works well when your site has traffic and data in Search Console. On a new site, or a site with few impressions, the list will be short and the data unstable. A few more things worth knowing:

  • Average position is an average. A page at position 12 may sometimes show up at 6 and sometimes at 25, depending on device, location and timing. That’s why it’s worth working with a date range of several months.
  • Sometimes the problem is the site, not the page. If every page on the site is stuck in the same range, it may be the site’s overall authority. The prompt is required to say so explicitly, and in that case the fix isn’t on a single page.
  • Claude can’t see your pages. The Search Console export contains numbers, not content. Unless you paste in the page, part of the diagnosis is inferred from the URL and keyword alone, and the prompt flags this.
  • There’s no guarantee of page one. The outcome depends on the competition, the quality of the change and the site as a whole. The diagnosis shows you where it’s most worth starting, not when the page will move up.

And if along the way you find weak pages that bring in nothing and aren’t close to ranking anywhere, those belong to a different process. We covered it in our guide Content pruning with a prompt that decides what to merge, upgrade or remove.

FAQ

Why an average position between 8 and 20 specifically?

Because that’s the range covering the bottom of page one and page two. Google already considers these pages relevant, so a relatively small change can move them, unlike pages at position 50 that need much more work.

How do you filter by average position in Search Console?

In the Performance report, turn on the Average position metric, then use the table’s row filter to keep only positions greater than 8. If you can’t also add a less-than-20 condition, export the data and filter it in a spreadsheet.

What if I don’t have competitor URLs?

The prompt also works without them, using only the keywords and data. It will state at the top of its answer that the diagnosis is less reliable, because without a competitor comparison it’s hard to tell what’s missing in coverage and search intent.

How long does it take to see results after a fix?

There’s no way to know in advance. It depends on when Google recrawls the page, the competition and the size of the change. That’s why the prompt forbids Claude from promising timelines, and it orders the work starting from the cheapest change.

Should I change the title of a page that already drives traffic?

Only with care. If the page ranks well for other keywords, changing its title or URL can hurt those rankings. First check which keywords it drives traffic from, and only then decide.

Want us to diagnose your pages stuck on page two?

  • Analysis of your current organic performance
  • 3 immediate opportunities identified
  • No commitment

Book a call →

Share this article: WhatsApp LinkedIn Facebook X

Want us to do this for you?

DigiAssets helps businesses rank on Google and get cited by AI engines like ChatGPT and Gemini. No generic packages, and reporting that ties every action to a result.

Skip to content