Signal-Based Cold Email (2026 benchmarks)
SkillCommunicationWrite a signal-based cold outreach email that earns a reply. Use when drafting outbound/SDR email to a prospect with a known recent signal (funding, hiring, exec hire, launch, community post).
Available today. Use it from your connected AI after setup.
No other account needed.
Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
Then ask your AI: use the Signal-Based Cold Email (2026 benchmarks) skill
What this skill tells your AI
The instructions your AI receives, as published by kami-community/kami in skills/signal_cold_email/SKILL.md and read by ahel’s review.
Baseline cold reply rate is ~3.4%. Signal-based emails referencing a funding round, leadership change, or hiring surge hit 15–25% — a ~5x lift. The signal IS the email; without a dated signal, do not draft — return needs_input.
Structure (exactly this order)
- Hook — why you are writing, tied to the dated signal. Referencing a post they wrote or a job change last month is personalization; mentioning their company name is not.
- One-sentence value — what you do for someone in their exact situation.
- One proof point — a single number or named customer. One, not five.
- Single low-friction CTA — one question or one 15/20-min ask. Never two asks.
Hard rules (Reviewer will reject violations)
- Plain text. 3–5 sentences. Under 150 words.
- Exactly one CTA.
- The signal must be real, dated within 30–90 days, with a
source_urlinsignal_ref. - Subject line under 50–60 chars, thesis-fit, no clickbait, no ALL CAPS.
- Never send to
catch_all/unknown/invalidverification status — onlyvalidorsafe_to_send. - Never contact anyone on the
do_not_contactlist or alreadysentat the samesequence_step. - Prefer a verified public/role inbox scraped from the company site (or founder-provided email). Never invent an address.
- Prioritize recipients by Fit × Intent (high fit + fresh signal first). Personalize from the dated signal and anything verifiable about the person online — relevance beats clever openers.
Follow-up sequencing (real-agency cadence)
- Sequence length 4 emails max (first email captures ~58% of replies; diminishing after 4).
- Cadence: day 0 → day 3 → day 7 → day 14. Set
followup_ataccordingly when a send completes. - Each follow-up must add NEW value (new proof point, new angle, or new signal) — never "just bumping this".
- Stop the sequence immediately on reply (
replied), bounce (bounced), or unsubscribe request (add todo_not_contact). - Respect
business_rules.max_sends_per_day(default 35/mailbox).
Output
Return a Draft object with subject, body, cta, signal_ref, sequence_step, and a filled self_check (word_count, cta_count, has_signal). Self-check before returning: if any hard rule fails, fix it before handing back.
Signals
- GitHub stars
- 64
- Forks
- 9
- Last commit
- Aug 2026
Advanced
- Item type
- skill
- Key
signal-cold-email- Source
- github.com/kami-community/kami