Newsletter & Digest Emails

⬅️ Day 6: Grow Overview

If you're building a B2B SaaS in 2026 with content (activity, notifications, analytics, recommendations), users often prefer digest emails over instant notification firehoses. The naive approach: send emails on every event. The structured approach: digest patterns (daily / weekly / monthly summaries), preference management, dynamic content (personalized per user), template system, deliverability discipline, unsubscribe compliance, A/B testing. Done well, digests drive retention + re-engagement; done poorly, they're spam-folder fodder. (Distinct from onboarding-email-sequence-chat (drip campaigns) and notification-preferences-unsubscribe-chat (broader preferences).)

1. Decide digest type — daily, weekly, monthly, real-time

Pick digest cadence + type.

Cadence options:

Real-time per-event:
- Email per significant event
- High volume; can spam
- Best for: critical notifications (security, billing)
- Worst for: activity feeds (overload)

Daily digest:
- One email per day with day's events
- Best for: active users; collaboration tools
- Examples: Linear daily digest, GitHub daily digest, Slack daily summary

Weekly digest:
- One email per week with week's activity
- Best for: less-frequent engagement; content / insights
- Examples: Substack reader digest, LinkedIn weekly, GitHub weekly stars

Monthly digest:
- Stats + recap
- Best for: dashboards, periodic insights
- Examples: Spotify monthly listening, Stripe revenue summary

Quarterly / annual recap:
- Year-in-review (Spotify Wrapped style)
- High engagement; viral potential

Per-user customization:

Cadence preference:
- User picks: real-time / daily / weekly / off
- Per-event-type granularity

Default for new users:
- Weekly typically (lower friction)
- Onboarding email teaches benefits

Mute / pause:
- "Pause for 2 weeks" option
- Vacation handling

For [USE CASE], output:
1. Cadence options
2. Default
3. Per-user preference UX
4. Mute / pause
5. Re-engagement after pause

The default-weekly rule: too aggressive (daily) for many users; weekly is safer default. Power users opt into daily.

2. Personalization — content per user

Generic newsletters get unsubscribed. Personalize.

Personalize digest content.

Personalization layers:

User-specific events:
- Activity in their workspace
- Items they own
- Items they're @mentioned in

Recommendations:
- Based on usage patterns
- Collaborative filtering
- "People similar to you read..."

Stats:
- Their numbers (activity / metrics)
- Comparison to peers (anonymized)
- Goals progress

Generic content:
- Newsletter / blog highlights
- Same for all users
- OK to include with personal

Implementation:

Server-side rendering:
- Build email per user
- Query user's data
- Inject into template
- Send via email provider

Email provider templating:
- Postmark, SendGrid, Resend support template + dynamic variables
- Pre-render server-side OR template-side

For high volume (>10K/day):
- Batch render
- Queue + worker
- Don't block API on send

Dynamic content blocks:

Render or skip:
- "If has new comments: show comments section"
- "If no activity: show recommendations"
- Don't send empty digest

Caching:
- Cache user data 5-15 min
- Reduce per-email queries

Privacy:
- Don't include sensitive data in email body
- Link to product for sensitive info
- Consider: GDPR DPA (email is processing)

Anti-patterns:
- Same content for all (boring)
- Stale data (user already saw)
- Sensitive info in email (security risk)

Output:
1. Personalization layers
2. Rendering pipeline
3. Caching
4. Privacy
5. Skip-if-empty rule

The "no empty digest" rule: don't email "you had 0 activities this week" — feels useless. Skip the send; or include recommendations / stats to fill.

3. Email templates — design + responsive

Email design is harder than web.

Design email templates.

Constraints:

Email clients:
- Gmail (best CSS support)
- Outlook (worst; strips many features)
- Apple Mail (good)
- Yahoo, others

Use:
- Tables for layout (not modern divs)
- Inline CSS (not external stylesheets)
- Web-safe fonts
- Inline images via CDN (not embedded)

Tools:

MJML:
- Markup language for emails
- Compiles to email-client-compatible HTML
- Used by: Mailjet, many startups

React Email (Resend):
- React components for email
- Compiled to HTML
- Modern dev experience
- Used by: 2026 default

Maizzle:
- Tailwind-based emails
- Component-friendly

Postmark / Sendgrid templates:
- Drag-drop designers
- Less flexible

Template structure:

Header:
- Logo
- Brand colors
- Unsubscribe link visible (legal)

Body:
- Personalized content
- Sections per content type
- CTAs to product

Footer:
- Unsubscribe (large, visible)
- Address (CAN-SPAM)
- Legal / preferences link

Responsive:
- Mobile-first design
- 600px max-width typical
- Stackable sections

Dark mode:
- prefers-color-scheme not always supported
- Some clients invert colors
- Test major clients

Plain text fallback:
- Always include
- Some clients prefer plain
- SEO / spam-folder benefit

Output:
1. Tool choice (React Email recommended)
2. Template structure
3. Mobile / desktop responsive
4. Dark mode handling
5. Plain text fallback

The 2026 React Email + Resend stack: modern dev experience for emails; deliverability via Resend. Replaces older MJML + SendGrid for many.

4. Email content + UX

What goes in the email?

Design email content.

Content blocks:

Greeting:
- "Hi [name]"
- Personalize with first name
- Skip if no name; use generic

Hero / TL;DR:
- "Here's your week" or similar
- Single most-important takeaway
- Primary CTA

Activity / events section:
- Bulleted list or cards
- Avatars / icons
- Link to detail in product

Stats section:
- Visual numbers
- Trends (up/down)
- Comparisons

Recommendations:
- Personalized suggestions
- "You might like..."

Featured content:
- Newsletter article
- New product feature
- Case study

CTA section:
- Primary call-to-action
- Single, prominent
- Secondary CTAs minimal

Footer:
- Unsubscribe (required)
- Manage preferences
- Help / contact

Length:

Sweet spot: 200-400 words
- Long: lose attention
- Short: not enough value
- Visual elements help

Subject line:

Length:
- 30-60 chars (mobile preview cuts ~30)

Patterns:
- "[Product] weekly: [Highlight]"
- "Your week at [Product]"
- "[Specific number]: [event]"
- Personalize when possible

Avoid:
- Spammy ("Free!" "Buy now!")
- All caps
- Excessive emojis (1-2 OK)

Preview text:
- Hidden text after subject
- 50-100 chars
- Complements subject

Image-to-text ratio:
- More text than images (deliverability)
- Images alt-text always

Output:
1. Content block library
2. Subject + preview line
3. Length guidelines
4. Image / text balance
5. Tone consistency

The "TL;DR at top" pattern: many users skim subject line + first line. Put the value upfront; details below.

5. Deliverability — sender reputation matters

Without good deliverability, emails go to spam.

Optimize deliverability.

Authentication:

SPF (Sender Policy Framework):
- DNS record listing authorized senders
- Required

DKIM (DomainKeys Identified Mail):
- Cryptographic signature
- Required

DMARC:
- Policy for SPF/DKIM failures
- Recommended; set to "quarantine" or "reject"

BIMI (Brand Indicators):
- Show logo in inbox (Gmail, Apple)
- Requires DMARC + verified mark

Sender warming:

New sending domain:
- Start low volume (100s/day)
- Gradually increase
- Don't blast 100K from cold domain

Subdomain strategy:
- Send transactional from app.yourdomain.com
- Marketing from marketing.yourdomain.com
- Isolates reputation

Engagement-based reputation:

Inbox providers track:
- Open rate (proxy for "good email")
- Spam reports (kills reputation)
- Replies (positive signal)
- Forwards (positive)

Improve:
- Send only to engaged users
- Sunset disengaged (no email open in 6 months)
- Re-engagement campaign before sunset

Content:

Spam triggers:
- Excessive caps
- Multiple ! in subject
- "Free", "Buy now", money symbols
- Mismatched From / Reply-To
- Heavy images, low text

Test before send:
- Mail Tester (mail-tester.com) score
- Spam check (GlockApps, Litmus)
- Inbox preview (multiple clients)

Bounces:

Hard bounce: invalid email
- Suppress immediately
- Don't retry

Soft bounce: temporary
- Retry 3x
- Suppress after 5 soft bounces

Suppression list:
- Hard bounces
- Spam complaints
- Manual unsubscribes
- Don't email forever

Tools:

Email providers:
- Postmark (transactional + marketing; great deliverability)
- Resend (modern; React Email integration)
- SendGrid (legacy; volume)
- Mailgun (developer-friendly)
- AWS SES (cheap; DIY)

Output:
1. Authentication (SPF / DKIM / DMARC / BIMI)
2. Subdomain strategy
3. Sender warming
4. Engagement tracking
5. Bounce / suppression handling
6. Provider recommendation

The Postmark / Resend default for transactional: best deliverability; reasonable pricing. Don't use marketing tools (Mailchimp) for transactional emails — different reputation pools.

6. Preference management — give users control

Preferences UI matters for trust + compliance.

Build preference center.

Page: /settings/notifications/email

Sections:

Per-type preferences:
- Activity digest (daily / weekly / off)
- Mentions (instant / digest / off)
- Comments (instant / digest / off)
- Product updates (newsletter; on / off)
- Marketing (on / off)

Granular:
- Per-channel-per-type
- Email + push + in-app per event

Mute scheduling:
- Pause for 1 week / 2 weeks / 1 month
- Vacation auto-resume

Unsubscribe:

One-click unsubscribe:
- CAN-SPAM requires
- Link in every email
- Single click → confirmation
- No login required (or auto-login)

List-Unsubscribe header:
- RFC standard
- Gmail / Outlook show "unsubscribe" button
- Required for inbox-promotions; nice-to-have for primary

Categories:
- Unsubscribe specific OR all
- Per-category checkboxes
- "Unsubscribe from all" option

Re-subscribe:
- User can resubscribe later
- Confirmation flow

Compliance:

CAN-SPAM (US):
- Clear unsubscribe required
- Visible business address
- No deceptive subject lines
- Honor unsubscribe within 10 days

GDPR (EU):
- Explicit consent for marketing
- Right to delete
- Data portability

CASL (Canada):
- Express consent required
- Explicit timestamps + records

CCPA (California):
- Similar to GDPR for CA residents

Implementation:

Suppression list:
- Store unsubscribed users / emails
- Check before every send
- Honor across all email types (or per-category)

Audit log:
- Who unsubscribed when
- For compliance investigations

Output:
1. Preference page
2. Unsubscribe flow
3. Compliance checklist (CAN-SPAM / GDPR / CASL)
4. Suppression list
5. Audit log

The one-click unsubscribe rule: required by Gmail / Yahoo as of 2024. List-Unsubscribe header + one-click endpoint. Without it, deliverability tanks.

7. A/B testing + measuring

Optimize newsletters.

A/B test digests.

Variables to test:

Subject line:
- "Your week at Product" vs "5 things at Product this week"
- Curiosity vs specific
- Length

Send time:
- Tuesday 9am vs Wednesday 9am
- Time zone (per-user vs sender's)

From name:
- Brand vs founder name
- "Product Team" vs "Sarah from Product"

Content:
- Long vs short
- Image-heavy vs text
- CTA placement

Testing:

A/B at scale:
- 10% to A; 10% to B; 80% to winner after 24h
- Need >10K subscribers for meaningful

Statistical:
- Use email provider A/B (Mailchimp / Resend)
- Or: track in product analytics

Metrics:

Open rate:
- Industry: 15-30%
- Higher = good subject + reputation
- Beware: Mail Privacy Protection (Apple) inflates

Click-through rate:
- 1-5% typical
- Higher = compelling content

Click-to-open ratio:
- CTR / open rate
- Better signal of engagement

Conversion:
- Click → in-product action
- Tied to your goal

Unsubscribe rate:
- <0.5% per send is healthy
- Spike = problem with content

Spam reports:
- <0.1% target
- Higher = reputation damage

Iterate:

Monthly review:
- Win-by-metric report
- Cumulative learnings
- Adjust template / cadence / content

Long-term:
- Cohort analysis
- Engagement-decay over time
- Sunset disengaged

Output:
1. Test framework
2. Variable matrix
3. Statistical approach
4. Metrics tracking
5. Iteration cadence

The "open rate inflation since iOS 15" reality: Apple's Mail Privacy Protection makes opens unreliable. Track click-through rate instead.

8. Sunset / re-engage disengaged users

Don't email forever to people who don't engage.

Sunset disengaged users.

Tracking:

Last engagement:
- Last open
- Last click
- Last in-product login

Sunset trigger:
- 90 days no email engagement
- Or: 180 days no in-product activity
- Configurable

Re-engagement campaign:

Day 1: "We've missed you"
- Highlight recent product changes
- Single clear CTA

Day 7: "One more thing"
- Different angle (success stories?)
- Easy unsubscribe

Day 14: "Last email"
- "We're going to stop emailing"
- One last value prop

If still no engagement:
- Move to suppression list
- Or: degraded list (very rare emails only)
- Or: hard unsubscribe

Win-back campaign (optional):

After 6 months suppressed:
- "Come back" with offer
- One-time outreach
- Low expectation

Anti-patterns:

Email forever:
- Tanks deliverability
- Disrespects user
- Anti-pattern

Aggressive re-engage:
- Multiple emails per day
- Spam complaints
- Worse than sunset

Sunset only on hard bounce:
- Misses disengaged
- Reputation slowly decays

Output:
1. Sunset trigger
2. Re-engagement sequence
3. Suppression after
4. Win-back rules
5. Tracking

The sunset discipline: hard but necessary. Better to email 50K engaged users than 200K + 150K disengaged. Inbox providers reward engagement-led senders.

9. Transactional vs marketing distinction

Two very different email categories.

Distinguish transactional vs marketing.

Transactional:

Purpose:
- User-initiated event triggered email
- Examples: receipt, password reset, verification, invitation accepted

Compliance:
- CAN-SPAM doesn't require unsubscribe (debatable)
- GDPR: legitimate interest basis
- Can send to unsubscribed-from-marketing users

Volume:
- Per-user per-event
- Typically lower volume

Provider:
- Postmark, Resend, SendGrid, AWS SES
- Separate pool from marketing

Marketing:

Purpose:
- Engagement / promotion / education
- Examples: newsletter, product updates, sales

Compliance:
- Explicit unsubscribe required
- GDPR: explicit consent
- Honor preferences

Volume:
- Per-list per-campaign
- Higher volume per send

Provider:
- Mailchimp, Klaviyo, Iterable, ConvertKit, Beehiiv
- Different reputation

Mixed (digest):

Where do digests fit?
- Activity digest of user's content = transactional-ish
- Newsletter = marketing
- Be honest; don't sneak marketing into transactional

Best practice:

Separate domains / subdomains:
- transactional.yourdomain.com (Postmark / Resend)
- marketing.yourdomain.com (Mailchimp / Klaviyo)
- Isolated reputation

Don't mix:
- Marketing email pretending to be transactional
- Transactional with marketing add-ons
- Risk: complaints; reputation damage

Output:
1. Categorization framework
2. Provider per category
3. Subdomain strategy
4. Compliance rules
5. Mixed-content handling

The mistake: sending marketing newsletter from transactional service. Reputation hit; eventually marked as spam.

10. Production checklist

Pre-send checklist.

Technical:

- [ ] SPF / DKIM / DMARC configured
- [ ] BIMI for brand visibility
- [ ] DNS records propagated
- [ ] Suppression list checked
- [ ] Bounce handling configured
- [ ] Test send to internal list
- [ ] Mobile + desktop preview tested
- [ ] All major clients tested (Gmail / Outlook / Apple)
- [ ] Plain text version included
- [ ] Image alt-text
- [ ] All CTAs link correctly
- [ ] Unsubscribe link works
- [ ] Mail Tester score >9/10

Content:

- [ ] Subject line tested for spam triggers
- [ ] Personalization tokens replaced (no "Hi {{first_name}}")
- [ ] Sensitive data not in body
- [ ] Brand voice consistent
- [ ] Legal compliance (CAN-SPAM / GDPR)
- [ ] Address in footer
- [ ] Unsubscribe visible

Process:

- [ ] Send to internal team first
- [ ] Hold for 1 hour for issues
- [ ] Monitor first batch
- [ ] Track opens / bounces in real-time
- [ ] Have rollback / pause-send capability

Output:
1. Pre-send checklist
2. Internal review process
3. Monitoring during send
4. Pause / rollback capability
5. Post-send analysis

The "send to yourself first" rule: every single send. Catches typos, broken links, bad rendering.

What Done Looks Like

A v1 newsletter / digest system:

Add later when product is mature:

The mistake to avoid: send to disengaged users forever. Tanks deliverability + signals desperation.

The second mistake: mix marketing into transactional emails. Different rules; mixing ruins reputation.

The third mistake: no preview / test before send. Mass-send broken email = mass-disengagement.