← ClaudeAtlas

fin-prospectinglisted

Defines ICP from top-customer cohorts, builds scored target lists, mixes outbound channels with experiments, and computes the activity math to hit pipeline goals. Use when user says "prospecting", "ICP", "target list", or .
meshcode-ai/skills-exec · ★ 0 · Testing & QA · score 63
Install: claude install-skill meshcode-ai/skills-exec
# Prospecting ## ICP — reverse-engineered from top customers Better than a good copy line, the ICP is **the common attributes of your top cohort**: industry · size · tech stack, plus **behavioral signals** (job postings, new funding, leadership changes, competitor defection). Attributes (fit) alone can't time the buying window. ## Account score = Fit 60 + Timing 40 Sort the list by numeric score instead of qualitative grades, and focus outreach only on the top 20%. Leave lower accounts in nurture channels. List accuracy beats volume — lists with a bounce rate above 5% are discarded immediately. ## Channel mix Allocate cold email · phone · LinkedIn · referrals · communities as monthly experiments (2 weeks per channel, same message). Record contact→opportunity conversion per channel and change next month's allocation accordingly. ## Activity math Target pipeline ÷ (contact→opportunity rate × average opportunity value) = required weekly contacts. Quotas are back-solved from activity volume — "work harder" is not a plan. ## Quality judgment A human reviews 10 of the top-20% accounts to check typos and personalization feasibility. For messages with under 3% response rate, check the message before the list (subject A/B, first-sentence value proposition). Re-entry deadlines are managed automatically by channel refresh cycles (email 6 months, phone 3 months). ## Output ICP definition (attributes · signals), scoring criteria table, weekly activity targets per channel, list quali