early-access-email-sequence（早期访问电子邮件序列）

Early Access Email Sequence Generator

Research a new Goose signup, then generate all 7 onboarding emails personalized with real context about their role and company. Output goes to a shared Notion database — one row per person, every email ready to copy-paste and send on schedule.

The sequence runs over 14 days (Day 0, 2, 5, 8, 10, 12, 14). Emails 2 and 4 are personalized by role. Emails 1, 5, 6 are universal. Emails 3 and 7 branch by activation status (generate the default version; note when to swap).

Inputs

For batch processing, accept a list of {linkedin_url, company_domain, signup_date} objects and process each one sequentially.

Procedure

Phase 1: Research the Person

Use web search to gather context on the signup. Do NOT scrape LinkedIn directly — use web search queries that surface LinkedIn profile data.

1. Search for the person:

- WebSearch: "{{linkedin_url}}" OR site:linkedin.com/in/{{handle}} - WebSearch: "{{person_name}}" {{company}} LinkedIn (once you have their name)

1. Extract:

- Full name - Current title and role - Company they work at (confirm it matches the provided domain) - Seniority level (IC, Manager, Director, VP, C-suite, Founder) - Any recent LinkedIn posts or activity themes (if surfaced in search results) - Background: previous companies, career trajectory

1. If LinkedIn data is thin, supplement with:

- WebSearch: "{{person_name}}" {{company}} (company blog, press mentions, podcast appearances) - This additional context helps personalize Emails 2 and 4

Phase 2: Research the Company

Run these searches in parallel where possible:

1. What they do:

- WebFetch: https://{{company_domain}} — extract tagline, product description, target audience - If the site is JS-rendered and returns little content, fall back to WebSearch: {{company_domain}} what does {{company}} do

1. GTM motion:

- Check for self-serve signup / free trial (PLG signal) - Look for "Book a demo" / "Contact sales" (sales-led signal) - Check pricing page: transparent pricing = PLG, "Contact us" = sales-led - WebSearch: {{company}} careers sales OR marketing OR growth — are they hiring GTM roles?

1. Size and stage:

- WebSearch: {{company}} funding round series — latest funding, stage - WebSearch: {{company}} headcount employees — approximate team size - Check LinkedIn company page follower count if surfaced

1. Competitors:

- WebSearch: {{company}} competitors alternatives vs - WebSearch: {{company}} G2 category — what category are they in? - Identify 2-3 named competitors (used to personalize Email 4 marketing variant)

1. Compile a company summary (2-3 sentences):

- What they do, who they sell to, how they sell (PLG vs sales-led), approximate size, key competitors

Phase 3: Classify Role

Based on the person's title, classify into one of four buckets. This determines which variants of Emails 2 and 4 to use.

Ambiguous cases:

• "Head of Growth Marketing" → Marketing (has "Marketing" in title)
• "VP Revenue" → Sales (Revenue usually maps to sales leadership)
• "COO" at a 200-person company → Growth/RevOps (not a founder wearing all hats)
• "Product Manager" → Growth/RevOps (closest fit; PM cares about activation, not sales)
• If truly ambiguous, default to Founder — it's the most general variant

Phase 4: Generate Emails

Read the email templates from references/email-templates.md in this skill's directory.

For each email, apply the research to personalize:

Email 1 (Day 0) — Welcome:

• Replace {{first_name}}, {{sender_name}}, {{calendly_link}}
• No heavy personalization needed

Email 2 (Day 2) — Use Case #1:

• Select the variant matching the role bucket
• Replace all standard variables
• Write the [PERSONALIZE] section: 1-2 sentences connecting the workflow to something specific about their company, based on your Phase 2 research. This is where the value is — reference their GTM motion, their market, their competitors, or their stage. Be specific and concrete, not generic.

Email 3 (Day 5) — Nudge:

• Generate the "not activated" version (default)
• Add a note: "If {{first_name}} has completed a workflow by Day 5, swap to the activated variant: [include activated variant text]"

Email 4 (Day 8) — Use Case #2:

• Select the variant matching the role bucket
• Replace all standard variables
• Write the [PERSONALIZE] section: connect the second workflow to their specific situation. Use competitor names, industry events, or market context from your research.

Email 5 (Day 10) — Community Spotlight:

• Output as template with placeholders intact ({{top_workflow}}, {{X}}, etc.)
• Add a note: "WRITE THIS LIVE on {{send_date}} with real usage data. Do not send with placeholder content."

Email 6 (Day 12) — Skills Page:

• Replace {{first_name}} and {{sender_name}}
• No personalization needed

Email 7 (Day 14) — Feedback Ask:

• Generate the "not activated" version (default)
• Add a note: "If {{first_name}} has been active, swap to the activated variant: [include activated variant text]"

Phase 5: Calculate Send Dates

From the signup_date, calculate each email's send date:

Phase 6: Output to Notion

First-time setup (no existing database)

If notion_db_url is not provided, create a new Notion database. Use the Notion MCP notion-create-database or create a database page manually.

Database schema:

Adding a row

Use notion-create-pages with the database as parent. Fill all properties from the generated content.

In the Notes column, include:

• The activated variant text for Email 3 (to swap if needed)
• The activated variant text for Email 7 (to swap if needed)
• Any special observations from research that might be useful for the sender

After creating the row

Confirm to the user:

Added {{name}} ({{role_bucket}}) at {{company}} to the email sequence.
- Email 1 ready to send: {{email_1_send_date}}
- Email 2 ready to send: {{email_2_send_date}} ({{role_bucket}} variant: {{use_case_1}})
- ...through Email 7 on {{email_7_send_date}}
- Company summary: {{company_summary_short}}

Batch Processing

When given multiple signups at once:

1. Collect all {linkedin_url, company_domain, signup_date} entries
2. Process each person sequentially through Phases 1-5
3. Add all rows to the same Notion database
4. After all are processed, show a summary table:

Email Writing Rules

These are non-negotiable — every generated email must follow them:

1. No filler openers. Never "I hope this finds you well", "I hope you're having a great week"
2. Under 150 words per email. Most should be 80-120 words.
3. One CTA per email. Never two asks in one email.
4. Sound like a founder, not a marketing team. Casual, direct, no corporate speak.
5. The [PERSONALIZE] sections are the entire value. Generic personalization ("since you're in B2B SaaS...") is worthless. Reference specific things: their competitor names, their GTM motion, their funding stage, their market. If research didn't surface anything specific enough, say something honest about why this workflow matters for companies at their stage.
6. Sign off with name only. No "Best regards", no title, no company. Just the name.
7. No emoji in emails. Professional but warm.

Example

Input:

linkedin_url: https://www.linkedin.com/in/jamesmith
company_domain: posthog.com
signup_date: 2026-03-10

Research findings:

• Jane Smith, Head of Growth at PostHog
• PostHog: open-source product analytics, PLG motion, ~80 employees, Series B ($45M), competes with Amplitude, Mixpanel, Heap
• Role bucket: Growth / RevOps

Email 2 personalization (Growth/RevOps variant):

Since you're running growth at PostHog, this one's for you:... Given PostHog's open-source PLG motion and the Amplitude/Mixpanel competitive landscape, the audit's competitive positioning section will show you exactly where PostHog's content gaps are vs. those incumbents — and the white space map tends to surface channels that open-source companies underinvest in.

That level of specificity — naming their competitors, referencing their PLG motion, noting the open-source angle — is what makes the email feel personal rather than templated.

Dependencies

• Web search and web fetch capabilities (for research)
• Notion MCP tools (notion-create-pages, notion-fetch, optionally notion-create-database)
• Email templates at references/email-templates.md in this skill's directory