Email Nurture Planner

This skill helps you plan content-driven email nurture sequences for Tiger Data. It produces a strategic sequence plan — goals, audience mapping, email-by-email outlines with subject line directions, key messages, and CTAs — but stops short of writing full email copy. The output is a blueprint that a writer (or the brand-voice-writer skill) can execute against.

When to use this skill

Planning a new email drip campaign or nurture sequence
Designing an educational email series around a topic (e.g., "time-series best practices in 5 emails")
Mapping out a content-driven subscriber journey
Deciding email cadence, sequencing, and content progression for a nurture flow
When someone asks "what emails should we send to nurture [audience] around [topic]?"

When NOT to use this skill

Writing the actual email copy (use brand-voice-writer for that)
One-off email campaigns or announcements (those don't need sequence planning)
Transactional emails (password resets, invoices, etc.)

Future expansion: This skill currently focuses on content/education nurture sequences. Welcome sequences, product onboarding sequences, re-engagement flows, and post-trial win-back sequences are not yet covered but are natural extensions. If a user asks for one of these, note that the skill doesn't have a dedicated framework for it yet and offer to adapt the nurture arc as a starting point.

Step 0: Pre-flight check

Read REFERENCES.md from the plugin root and run the pre-flight check described there. Call list_marketing_references() to verify Tiger Den is reachable. If it fails or the tool is not found, STOP — do not continue. Follow the error handling in REFERENCES.md.

Also fetch the No Fly List before doing any work:

get_marketing_reference(slug: "no-fly-list")

This is a list of customers who cannot be publicly referenced. Load these names as a hard constraint: never include any No Fly List customer in any output — not as named examples, proof points, customer quotes, case study references, or any other mention. If the user requests content featuring or referencing a No Fly List customer, stop and inform them that this customer cannot be publicly referenced. If a No Fly List name appears in source material you are working from, omit it from all outputs.

Instructions

1. Load brand and audience context

Before planning anything, fetch context from Tiger Den so the sequence is grounded in who we are and who we're talking to.

Fetch both reference docs:

get_marketing_context(slugs: ["product-marketing-context", "brand-voice-guide"])

product-marketing-context — audience personas, pain points, positioning, competitive landscape, and proof points. This determines who the sequence targets and what messages will resonate.
brand-voice-guide — email-specific tone and structural guidance. The guide has a dedicated email section with rules for subject lines, openers, CTAs, and voice.

If Tiger Den is not connected, do not proceed — tell the user to run /setup. See REFERENCES.md in the plugin root for details and error handling.

Freshness check: After fetching each reference doc, check the "Last updated" date (usually near the top of the doc). If either doc was last updated more than 6 months ago, flag it to the user before citing specific metrics or proof points. Say something like: "The product-marketing-context doc was last updated [date]. Some metrics or proof points may be outdated. Want me to proceed with what's there, or should we verify the numbers first?" This matters most for product-marketing-context, which contains specific performance benchmarks and customer metrics that change over time. Brand-voice-guide changes less frequently, so a 6-month lag there is less risky, but still worth noting.

2. Clarify the sequence brief

Gather inputs through a guided, conversational Q&A — ask one question at a time, wait for the answer, then ask the next. Do not list all questions at once. If the user's initial message already answers some of these, skip those and start from the first unanswered one.

Ask the following in order:

Goal — What should recipients do after the sequence? (e.g., start a trial, attend a webinar, adopt a feature, read a pillar piece)
Audience — Which persona or segment are we targeting? (You'll map this to the personas in product-marketing-context once you have the answer.)
Topic or theme — What content territory does this sequence cover?
Entry trigger(s) — How do people enter the sequence? More than one trigger is fine — for example, a sequence might apply to both newsletter signups and webinar attendees. Ask the user to list all relevant entry triggers, then confirm whether the same sequence should serve all of them or whether separate sequences make more sense.
Length — How many emails? If the user isn't sure, propose 4–6 and ask them to confirm before continuing.
Cadence — How often should emails send? If the user isn't sure, propose every 3–5 days and ask them to confirm before continuing.
Sender and sequence mode — Who should these emails come from?
- Default: Marketing nurture. Sender is "The Tiger Data Team." Tone is educational, brand-voiced. Emails read like valuable content that showed up in your inbox. This is the standard mode.
- Variant: Sales-assist. If the entry trigger(s) suggest prior human interaction (e.g., "met at AWS Summit," "post-demo follow-up," "event attendees"), ask: "Should this be a marketing nurture sequence (sent from the brand) or a sales-assist sequence (provided to the sales team to send from their own names)?" Sales-assist sequences are still planned by marketing but designed to be sent by reps. They're shorter (3–4 emails), more direct, can reference the prior interaction ("Great meeting you at..."), and use a named sender. The tone stays helpful and educational, not pushy, but warmer and more personal than a brand email.
- If unsure whether the context calls for sales-assist, ask. Don't guess.

Only proceed to Step 3 once all seven inputs are confirmed.

Step 2.5: Check for existing brief in Tiger Den

Generate a deterministic working title from Step 2 inputs: "{topic} nurture for {audience}". Append "(sales-assist)" if the sequence mode is sales-assist. Example: "Time-series migration nurture for Senior DBAs (sales-assist)". This working title becomes the default sequence_name in Step 5 unless the user provides a better one.

Follow the Step A protocol in references/briefs-integration.md (from the plugin root):

Generate source_run_id and source_run_at.
Search for an existing brief using the working title and brief_type: "content".
If a match is found, ask the user whether to update it or create a new one.
Carry the brief_id forward if updating.

3. Design the sequence arc

Plan the sequence as a narrative arc, not a random collection of emails. Each email should build on the last and move the reader closer to the goal.

Use this progression framework for content/education nurture sequences:

Hook (Email 1): Validate the problem or curiosity that brought them in. Mirror the entry trigger. Establish what they'll learn across the sequence.
Foundation (Emails 2-3): Teach core concepts. Build understanding of the problem space. Use concrete examples and data points from our content library.
Application (Emails 3-4): Show how to apply the concepts. Link to tutorials, guides, or case studies. Start connecting ideas to Tiger Data's approach (without hard selling).
Proof (Email 4-5): Provide evidence — benchmarks, customer stories, comparisons. Make the case through results, not features.
Action (Final email): Clear, single CTA aligned with the sequence goal. Summarize the journey. Make the next step feel like a natural conclusion, not a sales push.

Not every sequence needs all five stages. A 3-email sequence might compress Foundation + Application into one email. Adapt the arc to the length.

Sales-assist variant: If the sequence mode is sales-assist (from Step 2), compress the arc. These sequences are shorter (3-4 emails), assume the reader already had a human touchpoint, and move faster to proof and action. A typical sales-assist arc: (1) Follow-up referencing the interaction + one valuable resource, (2) Proof or case study relevant to their use case, (3) Direct CTA (trial, docs, or book a call). Skip the extended foundation stage since the prior interaction already established context.

4. Plan each email

For each email in the sequence, produce:

Email number and role: Where it sits in the arc (e.g., "Email 2: Foundation")
Send timing: Days after entry trigger or previous email
Subject line direction: 2-3 candidate angles for the subject line (not final copy, just directions). Follow the brand voice guide's email subject line rules.
Preview text direction: A one-sentence direction for the preview text (~90-140 characters). It should extend or complement the subject line, not repeat it. This is the second line readers see in their inbox and directly affects open rates.
Key message: The one thing the reader should take away from this email (one sentence)
Target length: Approximate word count for the email body. Use 50-125 words for transactional/short emails, 150-300 words for educational content, 300-500 words for story-driven emails. This guides the writer during hand-off to brand-voice-writer.
Content outline: 3-5 bullet points covering what the email body should address. Do not reference the sequence itself in the outline — frame the content as if the email is arriving independently, not as an installment in a series.
CTA: What action the email drives and where it links
Supporting content: Existing Tiger Data content to reference or link to (blog posts, docs, case studies). If Tiger Den MCP is available, search for these; otherwise note what type of content would be ideal.

5. Add sequence metadata

After the per-email plans, include:

Sender: The "from" name for the sequence (carries from the brief; include here so it's visible in the plan and gets passed to brand-voice-writer during hand-off)
Sequence name: A working title for internal reference
Total emails: Count
Total duration: Days from first to last email
Primary CTA: The ultimate action the sequence drives toward
Exit conditions: When and why someone leaves the sequence early (e.g., "exits if they start a trial," "exits if they unsubscribe," "exits if they convert via another channel"). Define at least one positive exit (they converted) and one negative exit (they disengaged).
Success metrics: What to measure (open rates, click-through, conversion to goal action)
Branching notes: Any conditional logic suggestions (e.g., "if they click Email 3's CTA, skip Email 4 and go to Email 5")
Implementation platform: Note where the sequence will be built. This field is informational for whoever implements the sequence; it does not change the plan or copy. Defaults by sequence mode:
- Marketing nurture: HubSpot (Tiger Data's marketing automation platform)
- Sales-assist: Outreach (Tiger Data's sales engagement platform)
- If the user specifies a different tool, note it here instead.

6. Cross-check against brand voice

Review the plan against the email section of the brand voice guide:

No em dashes. The brand voice guide has an absolute rule: do not use em dashes in any content. Scan the entire plan (subject line directions, key messages, content outlines, metadata) and replace every em dash with a period, comma, colon, or separate sentence. This applies to the plan itself, not just final copy, because the plan feeds directly into brand-voice-writer.
Marketing tone, not sales tone. For standard marketing nurture sequences: check the plan for sales-sequence patterns and remove them. No "I noticed you..." or "I wanted to reach out" language, no references to "your conversation with" or "your account manager," no personalized sign-offs from a named rep. The tone should be educational and brand-voiced, like a smart blog post that arrived in your inbox, not a cold email or BDR follow-up. Exception: If the sequence mode is sales-assist, referencing a prior interaction and using a named sender is expected. But even sales-assist emails should stay helpful and educational, never pushy or generic-sales-y.
No series language in openers. Emails should feel like standalone, valuable messages — not installments in a drip campaign. Scan content outlines and any sample opener language for phrases like "in this series," "as part of this sequence," "in our last email," "this week's installment," or "you're receiving this because you signed up for X." Flag these and replace with directions that treat each email as its own thing. The reader should never be reminded they're in a funnel.
Are subject line directions following the rules? (No clickbait, no ALL CAPS, no false urgency)
Is the CTA progression natural? (Not every email should push a demo)
Does the sequence lead with problems and value, not features?
Is the cadence respectful? (Not too aggressive for the audience)

Flag anything that drifts from voice guidelines.

Output format

Present the plan directly in the conversation as structured text. Do not create a file, document, or report — just respond in chat. Use these sections:

Sequence overview (goal, audience, entry trigger, cadence, duration)
Sequence arc (visual summary of the progression)
Per-email plans (the detailed breakdown from Step 4)
Sequence metadata (from Step 5)
Voice check (results of the Step 6 cross-check: list any flags, or state "No issues found" if the plan passes all checks. Always confirm em dash compliance explicitly.)
Next steps (recommendations for execution)

Do not use Ghost Paper, markdown files, or any other document generation tool. The plan is a working artifact meant to be iterated on in conversation, not a deliverable.

Tiger Den content enrichment

These features use Tiger Den tools that are already connected (since you fetched reference docs in Step 1).

Content search

Use search_content to find existing Tiger Data content that each email could reference or link to. Search for topics related to each email's theme. This grounds the sequence in real content rather than hypothetical links.

Freshness filter: Only use content published within the last 18 months. Use the published_after parameter when calling search_content to filter results (e.g., published_after: "2024-09-01" if the current date is March 2026). Older content may reference deprecated features, outdated benchmarks, or retired product names. If the only results for a topic are older than 18 months, note the gap in the supporting content field (e.g., "No recent content available on this topic; consider creating a new piece") rather than linking stale material.

Voice profiles

If the user mentions a specific sender for the sequence (e.g., "these should come from Matty"), call get_voice_profile with that person's name to load their writing samples and voice notes. Use these to inform the subject line directions and tone notes in the plan.

Tiger Den brief persistence

After the voice check (Step 6) and output, follow the Step B protocol in references/briefs-integration.md (from the plugin root) to save the brief.

If the user re-enters the editing loop after a save, update the existing brief on the next exit from editing. The brief_id is already held from the initial create or from Step A.

Field mapping

Map the email nurture plan output sections to structured_fields as follows:

Output section	structured_fields key
Goal	`sequence_overview.goal`
Audience	`sequence_overview.audience`
Entry trigger(s)	`sequence_overview.entry_triggers`
Cadence	`sequence_overview.cadence`
Duration	`sequence_overview.duration_days`
Sequence mode	`sequence_overview.mode`
Sequence arc stages	`sequence_arc.stages`
Per-email plans	`email_plans[]` (one object per email)
Sender	`sequence_metadata.sender`
Sequence name	`sequence_metadata.sequence_name`
Total emails	`sequence_metadata.total_emails`
Total duration	`sequence_metadata.total_duration_days`
Primary CTA	`sequence_metadata.primary_cta`
Exit conditions	`sequence_metadata.exit_conditions`
Success metrics	`sequence_metadata.success_metrics`
Branching notes	`sequence_metadata.branching_notes`
Implementation platform	`sequence_metadata.implementation_platform`
Voice check results	`voice_check`

Hand-off

After the plan is complete, offer to:

Write the full email copy using the brand-voice-writer skill (one email at a time)
Adjust the sequence based on feedback
Plan a follow-up sequence for non-engagers or completers

Do not auto-trigger brand-voice-writer. The plan is meant to be reviewed and iterated on before anyone writes copy. Wait for the user to confirm the plan looks good.

Editing the plan before writing

The user may want to tweak individual email plans before copy gets written. There are two ways to do this:

Conversational edits. The user says what to change (e.g., "Move the three evaluation tips from Email 1 to Email 2" or "Make Email 4's CTA about booking a call instead of restarting a trial"). Rewrite the affected email plan(s) and confirm before proceeding.
Paste-back edits. The user copies an email plan from the output, edits it directly (in a text editor, note, or the chat input), and pastes the revised version back. Use the pasted version as the new plan for that email, replacing the original. This gives the user full control over the outline without having to describe every change in words.

When presenting the plan, include this note at the end of the Next steps section: "To edit any email plan before writing, you can tell me what to change or paste a revised version of the plan directly."

Writing the emails

When the user is ready to write, ask which email to start with (suggest Email 1). Then hand off to brand-voice-writer with this context for each email:

The sequence overview (goal, audience, sender, sequence mode)
That specific email's full plan (role, subject line directions, preview text direction, key message, target length, content outline, CTA, supporting content)
The email's position in the arc ("This is Email 2 of 5, the Foundation email. Email 1 covered X, and Email 3 will cover Y.")

Write one email at a time. After each draft, ask if the user wants to revise it or move to the next email. This keeps the feedback loop tight and avoids throwing away bulk copy.

email-nurture-planner