Recipes / Hiring for the problem you solve: turn job posts into first lines and a lemlist campaign

Hiring for the problem you solve: turn job posts into first lines and a lemlist campaign

PredictLeads + Apollo + lemlist

Who it's for

SDRs who want a real reason to reach out. A company hiring for the problem you solve is telling you it hurts. This finds job posts that mention your category's pain points, finds the people who own that problem, drafts openers that start from the job post, and adds them to a lemlist campaign after you approve. PredictLeads can be swapped for LeadMagic, Apollo for ZoomInfo or Cognism, and lemlist for Instantly or Smartlead.

What you get

Matching job posts, the people to contact and openers waiting for approval. Sample output: every company and person in it is fictional.

Accounts checked: 40. Job posts in the last 30 days that mention your pain points ("manual reconciliation", "spreadsheet forecasting", "month-end close"): 3.

| Company | Job post | Matching text | Posted |
|---|---|---|---|
| Fernhill Software (fictional) | Senior Revenue Accountant | "reduce manual reconciliation across billing systems" | Oct 2 |
| Larkspur Freightworks (fictional) | FP&A Manager | "move forecasting out of spreadsheets" | Sep 27 |
| Mossgate Labs (fictional) | Controller | "shorten month-end close from 10 days" | Oct 9 |

Apollo people search (no credits used):
- Fernhill Software: Dana Okafor, VP Finance
- Larkspur Freightworks: Sam Whitlock, Director of FP&A
- Mossgate Labs: Priya Raman, CFO

Opener for Dana (draft):
"Saw Fernhill is hiring a Senior Revenue Accountant to cut manual reconciliation across billing systems. Most teams we work with try to hire their way out of that first."

Opener for Sam (draft):
"Your FP&A Manager post says the goal is to move forecasting out of spreadsheets. That's usually a sign the model has outgrown the tool, not the team."

Waiting for approval:
1. Enrich 3 people in Apollo for work emails (uses Apollo credits).
2. Add 3 leads to lemlist campaign 'Finance ops: hiring signal'. Campaign status: paused. It will stay paused.
Nothing has been enriched or added yet.

Time and cost

Estimate About 20–30 minutes a week, most of it editing openers.

Assumptions: One weekly run over a list of about 40 target accounts, finding 3–5 matching job posts, with one person enriched and added per match after approval.

• PredictLeads: One jobs request per account checked, so about 40 requests. Each request counts toward your plan's monthly request limit. Every call uses credits. Each MCP request counts toward the plan's monthly request limit. Once the limit is reached, requests return HTTP 402. PredictLeads pricing: 1 API credit per API request; discovery endpoints (e.g. Similar Companies) are billed per returned company. Source: docs.predictleads.com, checked 2026-10-16
• Apollo: One free people search per matching company, then after you approve, one people enrichment per person for a work email, which uses credits. Some tools use credits. Company search, company enrichment (single and bulk), company job postings, and people enrichment (single and bulk) use credits; people/contact search, CRM, sequence, email, task and analytics actions do not. Get conversation insights uses credits only if the conversation has AI insights. Source: docs.apollo.io, checked 2026-10-16
• lemlist: A read of the campaign and its status, then after you approve, one lead added per person. No campaign start, no sends, and no lemlist enrichment. Some tools use credits. Enrichment tools (find/verify email, find phone, LinkedIn enrichment, enrich lead/list), Purchase domain, Provision mailbox and Call lemlist API are flagged as consuming lemlist credits; credit amounts depend on enrichment type and plan. Read tools are not flagged. Source: help.lemlist.com, checked 2026-10-16

Exact credit costs depend on your plan with each vendor. Check the tool pages before a large run.

MCP servers

Every server is confirmed in: ClaudeChatGPTClaude Code

• PredictLeads Not probed
  ClaudeChatGPTClaude CodeAccess not confirmedEvery call uses credits
• Apollo Up · auth
  ClaudeChatGPTClaude CodeCursorCopilotRead + writeSome tools use credits
  https://mcp.apollo.io/mcp
• lemlist Up · auth
  ClaudeChatGPTClaude CodeCursorRead + writeSome tools use credits
  https://app.lemlist.com/mcp

Set up in a chat app

Every server in this recipe is confirmed for Claude and ChatGPT. Connect each one, then paste the prompts below into one conversation.

• Connect PredictLeads without code (Claude, ChatGPT)
• Connect Apollo without code (Claude, ChatGPT)
• Connect lemlist without code (Claude, ChatGPT)

Developer setup

{
  "mcpServers": {
    "predictleads": {
      "url": "None"
    },
    "apollo": {
      "url": "https://mcp.apollo.io/mcp"
      // headless alternative: "headers": { "X-Api-Key": "YOUR_APOLLO_API_KEY" }
    },
    "lemlist": {
      "url": "https://app.lemlist.com/mcp"
    }
  }
}

Step-by-step prompts

Copy all prompts

1. Using PredictLeads, pull job openings from the last {days} days for these companies: {account list}. Flag any post whose title or description mentions {pain points}. For each match, show the company, job title, post date and the exact sentence that matched.
2. For each match I keep, search Apollo for people at that company with titles like {titles}. Use people search only, no enrichment. Show name, title and a short note on why they own this problem.
3. Write a two-sentence opener for each person where the first line quotes or describes their job post. Show me the text only. Then, only after I approve it, enrich the people I pick in Apollo for a work email. Tell me how many first.
4. Show me the status of my lemlist campaign {campaign}. If it's running, stop and tell me. If it's paused or not started, add the approved people as leads only after I approve it. Don't start the campaign or send anything.

Run it on a schedule (optional)

• Run it weekly. A job post is freshest in its first few weeks, before the role is filled.
• In a scheduled run, pull job posts and run the free Apollo search, then stop. Approve enrichment, openers and lemlist adds yourself.
• Keep the campaign paused and start it by hand in lemlist once you've read every lead and opener.

Where this goes wrong

• A keyword match that isn't really your problem. "Forecasting" in a demand-planning role isn't revenue forecasting. Quote the matching sentence from the post for every match, and drop any where the context doesn't fit.
• Old or filled job posts. Show the post date for every match and skip anything older than {days} days. An opener about a role that's already filled reads as lazy research.
• Leads added to a campaign that's already running. Read the campaign status before adding anyone. If it's running, stop and ask. New leads in a live campaign may start getting the sequence before you've read the openers.
• Credits spent on the wrong people. Run the free Apollo people search first and confirm the person owns the problem the post describes (often the hiring manager's boss). Enrich only the ones you pick.

Gotchas

• PredictLeads lists job openings among its tools, but its docs don't describe a keyword search across all companies. This recipe starts from your account list and reads each company's posts. Check what your PredictLeads plan returns before relying on it.
• Each PredictLeads MCP request counts toward your plan's monthly request limit. Once you hit it, requests fail with HTTP 402. Keep the account list to what you'll actually work.
• PredictLeads doesn't publish its MCP endpoint URL. Connect it from the Claude connectors directory or the ChatGPT plugins directory. On ChatGPT Enterprise or Edu, an admin may need to enable it first.
• Apollo people search is free. People enrichment uses credits, as do company search, company enrichment and company job postings. You don't need Apollo's job postings here since PredictLeads covers them.
• lemlist can start, pause and archive campaigns and send messages, so say plainly that it must not. Its enrichment tools (find or verify email, enrich lead) also use lemlist credits; skip them since Apollo finds the email.
• lemlist's MCP docs don't say how a per-lead opener is stored. Check that your campaign's first step pulls the opener from a lead field before you add leads, or paste openers in by hand.
• lemlist in ChatGPT web needs developer mode and a Business or Enterprise/Edu plan. On Claude Team or Enterprise, an Owner adds the lemlist connector first.

Download

Get the prompts, guardrails and expected output as one Markdown file. Paste it into a Claude Project or ChatGPT project's instructions, or save it as CLAUDE.md for Claude Code.

Download hiring-for-your-problem.md

Vendor sites: Apollo. If we join an affiliate program, these links will be labeled as affiliate links.