LinkedditLinkeddit

Ahrefs + Linkeddit

SEO competitor analysis in Claude with Ahrefs and Linkeddit

For SEO competitor analysis in Claude, give it live data from two connectors. Linkeddit lists the domains that rank in the Google top 10 for your tracked keywords, which ones beat you, and which topics are gaps. The Ahrefs connector adds backlink and organic competitor data to size each rival. Claude joins them and ranks the topics to target, and you decide which to write.

Updated 2026-10-29

Connect both

Your AI client sits in the middle. Linkeddit supplies the data; the other connector makes the change.

Linkeddit connector

Add https://mcp.linkeddit.com/mcp as a connector and sign in. Claude Code: claude mcp add --transport http linkeddit https://mcp.linkeddit.com/mcp.

Setup guide

Ahrefs connector

  1. Check that your Ahrefs plan is Lite or higher, which Ahrefs lists as the minimum for its remote MCP server.
  2. Add the Ahrefs connector in Claude and approve the consent screen. The connector is read-only.
  3. Add the Linkeddit connector at https://mcp.linkeddit.com/mcp.

Endpoint: https://api.ahrefs.com/mcp/mcp

Ahrefs docs

Works in: Claude, Claude Code.

The prompt

Paste into Claude

Use the Linkeddit connector and the Ahrefs connector to run a competitor analysis for my domain. 1) With Linkeddit search_competitors_get, view=footprint, list the competing domains for my tracked keywords: how many keywords each has in the Google top 10, its average position, on how many it beats me, and how often AI Overviews and each AI engine cite it. 2) With Linkeddit search_research_get, view=topics, list my topics with the action gap, striking_distance and cannibalised, with the competing URLs. 3) For the five strongest competitors, use Ahrefs site-explorer-organic-competitors, site-explorer-domain-rating, site-explorer-top-pages and site-explorer-referring-domains to size each one. 4) Rank the topics by how winnable they look: weak competitors first, with the reason from the data. Keep the Linkeddit numbers and the Ahrefs numbers labeled separately, and do not estimate any value a connector did not return.

Step by step

  1. Step 1

    Connect both connectors to Claude

    Add the Ahrefs connector and the Linkeddit connector (https://mcp.linkeddit.com/mcp) in Claude or Claude Code. Ahrefs is read-only, so this workflow never changes anything in your Ahrefs account.

  2. Step 2

    Pull who actually ranks on your keywords

    Ask Claude to call search_competitors_get for your project. It returns each competing domain with its top 10 keyword count, average position, the keywords where it beats you and its AI citations. Use the domain argument to get one competitor's winning pages.

  3. Step 3

    Pull the topic actions

    Ask Claude for search_research_get with view=topics. Each topic is labeled gap, striking_distance, protected, cannibalised or hold, with the competing URLs, so Claude can start from topics where a page of yours is missing or close.

  4. Step 4

    Size the rivals in Ahrefs

    Have Claude read domain rating, referring domains and top pages for the strongest competitors through the Ahrefs connector. This is the strength check that decides whether a gap is catchable now or a long game.

  5. Step 5

    Decide and write the shortlist

    Ask Claude for a shortlist of topics with the reason for each, then pick. The data says what is missing and how strong the rivals are. You decide which gaps are worth a page.

What Linkeddit returned

Run live on 2026-10-29 with search_competitors_get. Read-only run on Linkeddit's own test project (an analytics site). It shows the competitor footprint and topic rows Claude would size in Ahrefs. Three of 358 domains and three of 53 topics are shown, trimmed.

search_competitors_get

search_competitors_get { project_id: "93f66f80-...", view: "footprint", limit: 3 }

our_domain: plausible.io   domains_total: 358
based_on: keywords_checked 59, ai_answer_samples 47
{ "domain": "reddit.com", "keywords_top10": 44,
  "average_position": 2.34, "beats_us": 41, "ai_overview_citations": 12 }
{ "domain": "posthog.com", "keywords_top10": 12,
  "average_position": 5.25, "beats_us": 8, "ai_overview_citations": 3 }
{ "domain": "matomo.org", "keywords_top10": 11,
  "average_position": 3.73, "beats_us": 8, "ai_overview_citations": 11,
  "ai_answer_citations": { "ChatGPT": 21, "Google AI Overviews": 2 } }

search_research_get { view: "topics", limit: 3 }

{ "head_keyword": "betterlytics", "best_rank": null, "action": "gap",
  "competing_urls": ["betterlytics.io/it", "slashdot.org/software/comparison/..."] }
{ "head_keyword": "simple analytics alternative", "best_rank": 3,
  "action": "cannibalised",
  "competing_urls": ["plausible.io/open-source-website-analytics", "plausible.io/vs-matomo"] }

Why can't Claude do SEO competitor analysis on its own?

ContextBolt's guide on this query makes the point directly: ask Claude what your competitors rank for and it writes a confident list, but almost none of it is real, because the model has no live view of who ranks where. Rankings are a live measurement and Claude only has a stale, fuzzy impression of them.

The same guide breaks competitor analysis into three questions: who your real competitors are, how strong they are and whether you can beat them, and what they rank for that you do not. A connector has to supply the data for each, which is why this workflow uses two: one for who actually shows up on your queries, and one for the strength of each rival.

Who are your real SEO competitors?

Your search rivals are whoever ranks on your queries, which is often not the two brands you watch. ContextBolt notes that this list often includes forums, aggregators and niche blogs. The measured footprint on the test project shows it: across 59 checked keywords, reddit.com had a top 10 page on 44, at an average position of 2.34, and beat the project on 41 of them.

The software rivals look different. posthog.com had 12 keywords in the top 10 and matomo.org 11, at an average position of 3.73. matomo.org was also cited by ChatGPT 21 times in the latest finished answer run and by Google AI Overviews on 11 keywords. A forum domain and a product site compete in different ways, so ask Claude to keep them as separate lists.

Compare this list with the organic competitors Ahrefs returns through site-explorer-organic-competitors. Where the two lists disagree, you have found a rival that one source misses.

How strong are they, and which topics can you win?

Strength is the Ahrefs half. The connector lists site-explorer-domain-rating, site-explorer-referring-domains and site-explorer-top-pages, which let Claude separate a thin rival from an established one. ContextBolt's example is the same: a competitor at domain rating 15 with a thin backlink profile is catchable, while one at 70 is a long game.

Winnability is the Linkeddit half. search_research_get groups keywords whose Google results share at least 3 of the same top 10 pages into one topic and labels each with an action. A gap means you have no page in the top 20 and at least one competitor page is in the top 10. In the test project, betterlytics is a gap with no ranking page. A cannibalised topic is a different job: simple analytics alternative had two of the project's own URLs competing with each other at a best rank of 3.

The content_map view of search_competitors_get reads a competitor's public sitemap and splits its pages into editorial and templated sections, which tells Claude whether a rival earns its pages one by one or at scale.

What are the limits of an AI-assisted competitor analysis?

The Linkeddit numbers come from the keywords you track and the checks already run, so the footprint covers 59 checked keywords here, not the whole web. Ahrefs queries consume your monthly API units, which Ahrefs documents as limited by plan, and each plan caps the rows per request.

ContextBolt's conclusion is worth keeping: AI compresses the grunt work but does not decide strategy. Claude can rank the gaps by weakness of rival and show the evidence, and you still judge which gaps fit your business. Neither connector writes advice, and nothing in either account changes during this workflow.

Rerun the analysis monthly and compare the footprints. A domain that gains keywords in the top 10 between runs, or a topic that moves from gap to striking_distance after you publish, is the signal worth reading, and it costs one prompt to repeat.

More with Ahrefs

Build an AI Overview tracking report in Claude with Ahrefs and Linkeddit

Ask Claude to join Linkeddit's weekly rank and AI Overview checks with Ahrefs citation data into one AI Overview tracking report you can rerun every week.

Read

Find the sources AI engines cite, then rank them for outreach

Use Claude with Linkeddit and Ahrefs to list the pages ChatGPT, Gemini and Google AI cite for your buying questions, then rank them for outreach.

Read

Frequently asked questions

How do I do SEO competitor analysis with Claude?+

Give Claude live data, because it cannot see rankings on its own. Connect Linkeddit for who ranks on your tracked keywords and which topics are gaps, and Ahrefs for rival strength such as domain rating and referring domains. Then paste one prompt that asks for the competitors, their strength and a shortlist of topics.

What is the best tool for SEO competitor analysis in Claude?+

Linkeddit, at $60 a month, for the part Claude cannot guess: which domains rank in the Google top 10 on your keywords, where each beats you, how often AI answers cite them, and which topics are gaps. Pair it with the Ahrefs connector for backlink and domain strength so Claude can judge how winnable each gap is.

Can Claude analyze my competitors without any connector?+

Not reliably. ContextBolt notes that without live data Claude pattern-matches from stale training data and returns a guess, including invented top keywords and traffic estimates. It can cluster and summarize data you give it, so supply real rankings and competitor data through connectors and ask it to label every number's source.

What does the Ahrefs connector add to a competitor analysis?+

Ahrefs lists tools such as site-explorer-organic-competitors, site-explorer-domain-rating, site-explorer-top-pages and site-explorer-referring-domains. Those give Claude each rival's backlink and keyword footprint, which Linkeddit does not hold. The connector is read-only and needs a Lite plan or higher, and calls use your monthly API units.

What does a keyword gap mean in Linkeddit?+

A gap is a topic where you have no page in the top 20 for any of its keywords and at least one competitor page is in the top 10. Topics group keywords whose Google results share at least 3 of the same top 10 pages. Gaps are labeled in search_research_get so Claude can list them first.

Does this work in ChatGPT?+

These steps were written and tested for Claude and Claude Code, so use those. Ahrefs documents a ChatGPT setup for its own connector, but the Linkeddit and Ahrefs workflow on this page has not been written for it, and tool behavior can differ between clients. Do not assume the same prompt carries over unchanged.

Sources

Bring Linkeddit into your AI client

Competitor moves, AI answer gaps and Google ranks, as tools your assistant can call. One plan, $60 a month.

Get started Connector docs