# SPF Record
v=spf1 include:_spf.google.com include:sendgrid.net ~all

# DKIM Record
selector._domainkey.example.com IN TXT "v=DKIM1; k=rsa; p=MIGfMA0GCSqGSIb3..."

# DMARC Record
_dmarc.example.com IN TXT "v=DMARC1; p=quarantine; rua=mailto:dmarc@example.com"

Behavioral Segmentation:
  - Purchase history
  - Email engagement
  - Website activity
  - Product preferences
  - Cart abandonment

Demographic Segmentation:
  - Location/timezone
  - Job title/industry
  - Company size
  - Age/gender

Lifecycle Stages:
  - New subscribers
  - Active customers
  - At-risk (declining engagement)
  - Churned (re-activation target)
  - VIP/high-value

Day 0 - Welcome Email:
  trigger: subscription_confirmed
  content: Brand introduction, expectations
  cta: Complete profile

Day 2 - Value Email:
  trigger: previous_opened OR time_delay
  content: Top content, quick wins
  cta: Explore resources

Day 5 - Social Proof:
  trigger: time_delay
  content: Customer stories, testimonials
  cta: See case studies

Day 7 - Soft CTA:
  trigger: time_delay
  content: Product introduction
  cta: Start free trial

Hour 1 - Reminder:
  trigger: cart_abandoned
  content: Items in cart reminder
  cta: Complete purchase

Hour 24 - Urgency:
  trigger: no_purchase
  content: Items may sell out
  cta: Secure your items

Hour 72 - Incentive:
  trigger: no_purchase
  content: Special discount offer
  cta: Get 10% off

Subject Lines:
  - Length (short vs long)
  - Personalization
  - Emojis
  - Questions vs statements
  - Urgency words

Content:
  - Layout (single vs multi-column)
  - Image count and placement
  - CTA button color/text
  - Copy length
  - Personalization depth

Timing:
  - Send day
  - Send time
  - Timezone optimization

import scipy.stats as stats

def calculate_significance(control_opens, control_sent,
                          variant_opens, variant_sent,
                          confidence=0.95):
    """Calculate if A/B test result is significant."""

    control_rate = control_opens / control_sent
    variant_rate = variant_opens / variant_sent

    # Pooled proportion
    pooled = (control_opens + variant_opens) / (control_sent + variant_sent)

    # Standard error
    se = (pooled * (1 - pooled) * (1/control_sent + 1/variant_sent)) ** 0.5

    # Z-score
    z = (variant_rate - control_rate) / se

    # P-value
    p_value = 2 * (1 - stats.norm.cdf(abs(z)))

    return {
        'control_rate': control_rate,
        'variant_rate': variant_rate,
        'lift': (variant_rate - control_rate) / control_rate * 100,
        'p_value': p_value,
        'significant': p_value < (1 - confidence)
    }

def optimize_send_time(subscriber_data):
    """Analyze historical engagement to find optimal send times."""

    engagement_by_hour = {}

    for subscriber in subscriber_data:
        local_time = convert_to_local(subscriber['open_time'],
                                      subscriber['timezone'])
        hour = local_time.hour

        if hour not in engagement_by_hour:
            engagement_by_hour[hour] = {'opens': 0, 'total': 0}

        engagement_by_hour[hour]['opens'] += 1
        engagement_by_hour[hour]['total'] += 1

    # Calculate open rates by hour
    for hour, data in engagement_by_hour.items():
        data['rate'] = data['opens'] / data['total']

    # Find best hours
    sorted_hours = sorted(engagement_by_hour.items(),
                         key=lambda x: x[1]['rate'],
                         reverse=True)

    return sorted_hours[:3]  # Top 3 hours

-- Calculate subscriber engagement score
SELECT
    subscriber_id,
    email,
    COUNT(CASE WHEN event_type = 'open' THEN 1 END) as opens_30d,
    COUNT(CASE WHEN event_type = 'click' THEN 1 END) as clicks_30d,
    MAX(event_date) as last_activity,
    CASE
        WHEN COUNT(CASE WHEN event_type = 'open' THEN 1 END) >= 5 THEN 'highly_engaged'
        WHEN COUNT(CASE WHEN event_type = 'open' THEN 1 END) >= 2 THEN 'engaged'
        WHEN COUNT(CASE WHEN event_type = 'open' THEN 1 END) >= 1 THEN 'somewhat_engaged'
        ELSE 'unengaged'
    END as engagement_tier
FROM email_events
WHERE event_date >= CURRENT_DATE - INTERVAL '30 days'
GROUP BY subscriber_id, email;

Re-engagement Campaign:
  trigger: no_opens_60_days
  sequence:
    - Day 0: "We miss you" email
    - Day 7: "Last chance" with offer
    - Day 14: Final warning

  action_after_sequence:
    if: no_engagement
    then: move_to_suppression_list