12 Claude Prompts for B2B Founders

Copy-paste prompts for the six workflows founders run most: hiring signals, ICP enrichment, competitive intel, daily outreach, churn watch, and warm intros. Each uses LinkFetch + Claude via MCP.

By Fuat F, Founder, LinkFetch

Published

Updated

5 min read

These 12 prompts cover the six LinkedIn data workflows founders run most: weekly hiring signals, real-time enrichment, competitive monitoring, daily outreach, churn watch, and warm intros. Each runs in Claude Desktop with the LinkFetch MCP server. Copy the prompt, swap the bracketed values for your own watchlist and ICP, and run it. Estimated credit cost is noted for each. If you run a sales ops team rather than a founder-led GTM, the sales ops version of this library covers the same workflows tuned for a CRM-bound team. Founders who are actively raising will also recognise most of these workflows from the other side; the VC and dealflow prompt library shows what investors are running on the same signals.

How to use this library

Every prompt here uses LinkFetch's MCP tools: the jobs and company tools (search_jobs, search_companies, get_company), which run on LinkFetch's own dataset and need no LinkedIn account, and the people tools (get_person, search_people, get_person_posts), which run through the LinkedIn account you connect. To run any of these in Claude Desktop:

  1. Add the hosted LinkFetch MCP server to Claude Desktop (Settings, Connectors, add a custom connector) with the URL https://linkfetch.io/mcp?key=lf_your_key, using the API key from the MCP page of your dashboard. For the people prompts, connect your LinkedIn account under Accounts.
  2. Open a new conversation.
  3. Paste the prompt, fill in the bracketed placeholders, and run.

The prompts are written as direct instructions to Claude with tool calls embedded. Claude will call the LinkFetch tools automatically and synthesize the results into the output format described. The prompts that draft outreach also use Claude's own writing on the retrieved data; that part costs no LinkFetch credits.

Credit estimates assume a 50-account watchlist for the weekly monitoring prompts. Adjust by dividing or multiplying by the actual watchlist size. Keep the watchlist as a lead list in LinkFetch (Claude can save companies to one with save_lead) or paste it into the prompt.

Hiring signals (prompts 1 and 4)

Tracking new roles (prompt 1) is the highest-signal weekly workflow for founders selling to growth-stage companies. A new VP of Sales opening at a target account is not just a job post; it is a signal that budget was released for a function and the company is actively building. The prompt surfaces all such events across your full watchlist in a single Monday-morning run and ranks them by how closely the hire maps to your ICP's buying motion.

Finding accounts that just raised (prompt 4) is the fastest-converting outreach trigger available on LinkedIn. Founders who message within 48 hours of a funding announcement are reaching a company at peak openness to new vendor conversations. The prompt filters your watchlist to only accounts with a meaningful signal and drafts the opener in the same run, removing the research-to-outreach delay that lets the window close.

Teams using signal-based prospecting with Claude report 5 to 7x higher reply rates compared to untriggered cold outreach, with response rates on funding-trigger messages often exceeding 30% (lead411.com, 2026).

ICP and pipeline quality (prompts 2, 6, 10)

Enriching inbound signups (prompt 2) is the fastest ROI workflow in this library. A new sign-up arrives with a name and a work email. The prompt turns it into a structured ICP record (current title, company size, hiring signals, score) in under 30 seconds. The full ICP enrichment pipeline covers routing, CRM write-back, and edge cases in detail; this prompt is the condensed, single-shot version for founders doing it manually.

Scoring a prospect (prompt 6) standardizes qualification. Most founders carry an informal ICP in their heads and re-derive it for each prospect. A structured 10-point score, consistent across every evaluation, makes pipeline review conversations shorter and removes the variance that accumulates when different people informally qualify the same lead differently.

Pre-call prep (prompt 10) compresses the 20-minute LinkedIn research session before an intro call into a 30-second prompt run. The two suggested questions it generates are drawn from the actual company data, not generic BANT questions, but questions specific to what the data shows about what the company is working on right now.

Competitive intelligence (prompts 3 and 9)

Mapping a competitor's team (prompt 3) gives a hiring-based org chart that would otherwise require 45 minutes of manual profile browsing. The function-level breakdown tells you where a competitor is investing (a wave of engineering openings and none in sales is a product-led motion; the inverse is sales-led). Monitor it monthly and you will catch strategic shifts faster than you would from their blog posts or press coverage.

Tracking hiring delta across a peer set (prompt 9) is the most quotable output in this library. A table showing month-over-month hiring for your 10 closest competitors, updated monthly, is the kind of data that ends up in board decks, investor updates, and competitive battle cards. It takes under a minute to run.

Outreach (prompts 5, 7, 8, 12) and churn watch (prompt 5)

Daily outreach (prompt 7) is the 60-second version of the full daily outreach setup. It checks for new signals and drafts only the messages worth sending today, not a full queue, but a filtered shortlist. Run it at 08:30 and you review before 09:00.

Churn watch (prompt 5) catches account risk before it becomes a cancellation email. A customer whose champion just left, or whose company stopped hiring into the function that uses your product, is a churn signal. Catching it early gives you 2 to 3 weeks to engage the replacement champion or escalate to a contract conversation before the renewal arrives cold.

The signal-based founder outreach playbook covers the architecture behind prompts 1, 4, 7, and 12 in full, including watchlist construction, volume limits, and the review workflow that keeps message quality high without adding significant overhead.

FAQ

Do I need a paid Claude plan to run these?

Claude Pro ($20/month) is sufficient for personal use. Claude for Work (Teams plan) is the right tier if multiple people at your company will share the MCP integration. The LinkFetch MCP server is added once per user in Claude Desktop's connector settings; each user needs their own LinkFetch API key.

What is the fastest prompt to start with?

Prompt 6 (score a prospect against your ICP) requires no watchlist setup and produces immediate value on any live prospect. Run it on the last 5 accounts that came through your pipeline to calibrate the ICP criteria, then move to the weekly watchlist prompts once you have a list of 25+ target accounts.

How often should I update the prompts?

Update the ICP scoring criteria (prompt 6) whenever you close a new customer or lose a deal you expected to win; both events usually reveal something about your ICP definition. Update watchlist prompts (1, 4, 5) when you add or remove accounts. The outreach and pre-call prompts (7, 10, 12) are stable; adjust the tone if feedback from prospects suggests the copy is reading as templated.


Last updated: 2026-10-10 · Author: LinkFetch team

Try it on your own market with free credits.