I want personalized cold outreach for these 20 prospects

A ready-to-send cadence for a 20-prospect list, each message grounded in signal-based research.

A ready-to-send cadence for a 20-prospect list, every message grounded in actual signals rather than merge-field fluff.

The prompt

Paste this into Claude Code
$ Here are 20 prospect names and companies. Research them, score them against my ICP, and draft a personalized first-touch sequence.

What the Algorithm does

Bdr owns the ICP scoring and multi-channel sequence shape. Research does the enrichment pass — surfacing the buying signals that make each message feel written for one person. Sales wraps everything in stakes-based narrative so the outreach actually lands.

What you'll see
$ ♻︎ Entering the DOS ALGORITHM… (v0.0.1) ═════════════🗒️ TASK: Personalized cold outreach for 20 prospects━━━ 👁️ OBSERVE ━━━ 1/7🔎 REVERSE ENGINEERING:• Wants: 20 personalized sequences, signal-grounded• Not wanted: mail-merge slop, fake relevance💪🏼 EFFORT LEVEL: Extended | Per-prospect research plus drafting🏹 CAPABILITIES SELECTED:• Bdr (Score + Sequence) — ICP scoring, cadence shape• Research (Standard, per prospect) — signal enrichment• Sales — stakes-based narrative framing━━━ 📋 PLAN ━━━ 3/7Score list against ICP → enrich top 20 in parallel → draft sequences → narrative pass.━━━ ⚡ EXECUTE ━━━ 5/7Bdr scores across fit, signals, seniority, company size. Research pulls 2-3 fresh signals per prospect. Sales tightens the stakes line on each.✓ Done. 20 sequences drafted, 14 flagged as high-confidence.

What you get back

A campaign folder: one file per prospect containing the ICP score, the signals Research found, the full first-touch email, and a 3-step follow-up cadence. A summary sheet ranks the 20 by confidence and flags anyone below the ICP floor so you don't waste the first touch on bad fits.

Variations

  • Just score them, don't write the emails yet
  • LinkedIn-only sequences, skip email
  • Shorter sequences — 2 steps not 4
  • Use my existing templates as the starting point

Was this page helpful?