The viral version of this is a bot that blasts hundreds of generic applications while you sleep. Skip it: it gets your accounts flagged, and the data says it loses anyway. Personalized cold emails reply at around 17 to 18%, versus 7 to 9% for generic ones, roughly double, based on Woodpecker's analysis of over 20 million emails. So the winning move is fewer, better, researched emails. Claude just makes "researched" take minutes instead of hours.
Apify is a scraping platform, and it plugs into Claude as a connector. One-time setup, no code:
https://mcp.apify.comTell Claude your target and have it run a jobs scraper on Apify. The two most proven LinkedIn ones:
This is where most versions of this hack lie. A jobs scraper returns job postings, not a hiring manager's verified inbox. Getting a named person and a likely email is a separate second step, using an enrichment tool like an Apify people/email finder, or Hunter.io or Apollo.io (both have free tiers). And those emails are usually a best guess (pattern-matched like firstname@company.com), often not verified, so some bounce.
Claude researches each company and writes a personal first draft per role. You review, tweak, and press send yourself. No auto-send, ever, that's the line between a real outreach system and a spam cannon.
Once Apify is connected, paste this. Fill in the brackets:
You have access to my Apify connector. Help me run a focused, human-in-the-loop job hunt. Do NOT send anything, you research and draft, I review and send.
MY TARGET
- Role: [e.g. Senior Paid Media Manager]
- Location / remote: [e.g. Melbourne, or remote]
- Seniority: [e.g. mid-senior]
- Must-haves: [e.g. B2C, ad spend over $50k/mo]
- Deal-breakers: [e.g. agencies, on-site 5 days]
- About me (for tailoring): [3-4 lines: my edge, top results with numbers, what I want]
STEP 1, FIND ROLES
Run a LinkedIn jobs scraper on Apify (start with curious_coder/linkedin-jobs-scraper, or cheap_scraper/linkedin-job-scraper if keyword input is easier). Public postings only. Pull 30 to 50 matching roles. If LinkedIn input is fiddly, use an Indeed scraper instead. Tell me which Actor you used and roughly what it cost.
STEP 2, ORGANIZE
Return a clean table: Company, Role, Location, Posted date, Why it fits me (one line), Link. Rank by fit and flag the top 10.
STEP 3, RESEARCH (top 10 only)
Briefly research each company (what they do, recent news, likely challenge in my area). Note any named hiring contact only IF it appears in the posting or on their public careers page. Never fabricate a name or email, if you don't have a real one, say "apply via posting" and leave contact blank.
STEP 4, DRAFT (one email per role, for my review)
Write a short, genuinely personal email per top-10 role: 120 to 150 words, warm and specific, no corporate filler, no em dashes. Reference something real about that company, connect it to one concrete result of mine (with a number), and end on a soft question, not a hard pitch. Leave [NAME] as a placeholder if no real contact was found.
Show me everything as drafts. I will personalize further and send each one myself. Never auto-send.
๐ Find a real contact (only when you want it).
For these 5 companies where I have no named contact, use an Apify email-finder to find the likely hiring manager or lead and their best-guess business email. Label each clearly as VERIFIED or UNVERIFIED PATTERN-GUESS. For unverified ones, tell me so I can decide whether to use it, apply through the posting, or find the contact manually. Do not send anything.
โ๏ธ Sharpen one draft.
Take this draft and make it more specific to [Company]. Cut anything generic, add one line referencing [their recent thing], tie it to my result: [metric]. Under 140 words, no buzzwords, no em dashes. Give me two versions, one warm, one direct.
๐ Build a tracker.
Turn my top 10 into a tracker table: Company, Role, Contact, Email status (verified/guess/none), Draft ready?, Sent? (I'll fill), Follow-up date (+5 business days).
๐ค Claude: claude.com (add the connector under Customize, then Connectors)
๐ท๏ธ Apify: apify.com (the connector URL is https://mcp.apify.com)
๐ The reply-rate data: Woodpecker cold email statistics