feat(skills): add marketing skills collection

shared/skills/events/references/webinar-funnel.md

@@ -0,0 +1,88 @@
+# The Webinar Funnel — Registration → Show-Up → Live-to-Close → Nurture
+
+The webinar is the flagship hosted format because it's the whole event arc in miniature, repeatable monthly, and every stage is measurable. Work the four stages in order — each stage's conversion rate is a separate lever with separate fixes.
+
+**The funnel at a glance** (typical B2B ranges — directional benchmarks, not targets; your own trend line is the real baseline):
+
+| Stage | Metric | Typical range |
+|---|---|---|
+| Registration page | Visitor → registrant | 30–50% (warm traffic), 10–25% (cold) |
+| Show-up | Registrant → attendee | 35–45% live; lower for cold/ads traffic |
+| Hold | Attendee stays past minute 40 | 50–70% |
+| Convert | Attendee → next step (trial, demo, offer) | 5–15% of attendees for a soft CTA; 1–5% direct purchase |
+
+## Stage 0: Topic & Offer (decided before anything else)
+
+The topic does the targeting and most of the selling:
+
+- **Pick a problem-aware topic, not a product topic.** "How [ICP] does X without Y" out-registers "Intro to [Product]" — the audience you want shows up for their problem, not your roadmap.
+- **Name the transformation in the title**: specific outcome + specific audience + (optionally) a number or timeframe. Test titles the way you'd test ad headlines — the title *is* the ad.
+- **Decide the offer before writing the content.** What's the next step for an attendee who loved it — trial, demo, audit, purchase? The entire live structure builds toward that one step. A webinar with no decided offer becomes a lecture with an awkward ending.
+- One topic, one promise, one offer. Stack more and every rate drops.
+
+## Stage 1: Registration
+
+**The registration page** is a landing page (→ **copywriting** for craft); webinar-specific rules:
+
+- Headline = the promise from the title; subhead = who it's for and what they'll walk away able to do
+- 3–5 "you'll learn" bullets written as outcomes, not agenda items
+- Speaker credibility in one tight block (why should they listen to *you* on this)
+- Date/time with timezone handling; "can't make it? register anyway for the replay" — replay-registrants are real leads
+- Short form: name + email (+ one qualifying field max if sales needs it)
+
+**The promo plan** — start 2 weeks out, not 6 (urgency compresses better than it stretches):
+
+- **Email list** — 3 sends: announcement, value-add reminder (share a preview insight), last-call day-of (→ **emails**)
+- **Social** — founder/host personal posts outperform brand posts; share the *why this topic now* angle (→ **social**)
+- **Partners** — a co-hosted webinar doubles reach for free; partnership mechanics → **co-marketing**, but note: co-hosted registrant lists need explicit consent handling for both parties
+- **Paid** — only after the topic is proven organically; retargeting warm traffic to a reg page works, cold-to-webinar ads are an expensive way to buy no-shows (→ **ads**)
+- **Speakers' own audiences** — for panels/summits, every speaker promotes; make it effortless (pre-written posts, custom links)
+
+## Stage 2: Show-Up (the hardest metric)
+
+Registrants are cheap; attendance is the funnel's leakiest joint. The show-up system:
+
+- **Calendar add at registration** — the single highest-leverage fix. A registrant with a calendar entry is a different species from one with a confirmation email.
+- **Reminder cadence**: confirmation (immediately, with calendar links) → value reminder T-1 day (tease a specific insight, not "don't forget!") → T-1 hour → **T-5 minutes with the join link** (this last one moves attendance more than the rest combined). SMS reminders where consented lift show-up meaningfully (→ **sms**).
+- **Close the gap between registration and event.** Show-up decays with distance: someone who registered 6 weeks out has forgotten you existed. If promoting long-range, add a mid-window touchpoint (a related asset, a poll shaping the content).
+- **Pre-engagement**: ask a question at registration ("what's your biggest challenge with X?") — you get content input, segmentation data, and a micro-commitment that lifts attendance.
+- Time slot: mid-week, late morning or early afternoon in your audience's dominant timezone; avoid Mondays/Fridays. Test against your own data.
+
+## Stage 3: Live-to-Close (sell without being salesy)
+
+The arc that converts without feeling like a pitch:
+
+1. **Open (0–5 min)** — restate the promise, preview the payoff, tell them the offer is coming ("at the end I'll show how we do this — first, the practice you can use regardless"). Naming the pitch upfront *removes* the salesy feeling; the ambush is what people hate.
+2. **Content (5–35 min)** — teach the real thing. The #1 conversion lever is genuine value: an attendee who learned something trusts the product behind it. Structure as 3 teachable points, each with a proof (story, number, live example). Use attendee questions/polls to keep hold rate up.
+3. **The transition (1 min, scripted)** — the hardest 60 seconds; write it word for word. The honest bridge: "everything I showed you can be done manually — here's what it looks like when [product] does it for you." The product enters as the *implementation* of the content, not a topic change.
+4. **Offer (5–8 min)** — one offer, concretely: what they get, what it costs (or what the next step is), why now (a real reason — expiring bonus, cohort start, limited seats; never fake scarcity, → **offers** for legitimate urgency design).
+5. **Q&A (10+ min)** — conversion happens here; questions are objections in disguise. Seed 2–3 starter questions for cold starts, answer the objection behind the question, and re-state the offer + link once mid-Q&A and once at close.
+
+Hold-rate mechanics throughout: deliver on a specific promise made in minute 1 at minute ~35 (announced), use pattern breaks every ~7 minutes (poll, story, screen change), and never front-load housekeeping.
+
+## Stage 4: Post-Webinar (half the revenue is here)
+
+Segment by behavior, then sequence (→ **emails** for craft):
+
+| Segment | Play |
+|---|---|
+| **Attended, engaged** (stayed for offer, asked questions) | Personal follow-up within 24h referencing their question; direct next step |
+| **Attended, left early** | Replay + timestamp to what they missed; softer CTA |
+| **No-show** | "Sorry we missed you" + replay with a deadline. No-shows are warm — they raised their hand once; a 2–3 email replay sequence recovers a meaningful fraction |
+| **Replay-registrants** | Same as no-shows, minus the apology |
+
+- **Replay strategy**: time-limited replay (72h–1 week) preserves urgency for the offer; evergreen replay converts the offer to a standing CTA and becomes a lead magnet (→ **lead-magnets**). Pick per goal — limited for launches/offers, evergreen for education-led capture.
+- **Cart/offer close**: if the offer had a deadline, run a real close sequence (deadline reminder → objection email → final hours). All urgency claims must be true.
+- **Recycle the asset**: transcript → recap post (→ **content-strategy**), clips (→ **video**), quotable stats for AI-citable content (→ **ai-seo**). A monthly webinar run this way is a content engine with a lead-gen side effect.
+
+## Metrics That Diagnose
+
+- **Low registration** → topic/title/promise problem (or traffic quality). Fix the offer of the webinar itself before touching promo volume.
+- **Low show-up** (<30%) → reminder system or reg-to-event gap; check calendar-add rate first.
+- **Low hold** → content front-loading or promise mismatch; find the drop-off timestamp.
+- **High hold, low conversion** → transition or offer problem; the audience liked the class but wasn't shown a reason to act.
+- Cost per qualified attendee and per opportunity — comparable against your other channels, and the number that decides the program's future.
+
+---
+
+*Skill category identified via 2026-07 competitive research (webinar-marketing in alirezarezvani/claude-skills, MIT — idea credited; content authored from scratch to this repo's standard). Benchmarks are directional industry ranges — treat your own trend line as the baseline.*