Your bot is in EST. Prospect is in Sydney.
You send at 9am EST = 1am Sydney.

Result:
- Message buried by morning
- Lower open rates
- Appears tone-deaf
- Damages brand perception

"Why are they emailing me at 1am?"

Same scenario, timezone-aware:
You send at 9am Sydney time.

Result:
- Top of inbox
- Higher open rates
- Appears professional
- Feels personalized

"They actually know what time it is here."

Phone number: +61 2 9999 9999
+61 = Australia
02 = Sydney/NSW area code
Timezone: Australia/Sydney (AEST/AEDT)

Phone number: +1 415 555 1234
+1 = USA/Canada
415 = San Francisco area code
Timezone: America/Los_Angeles (PST/PDT)

Mapping:
- Extract country code
- For large countries (US, CA, AU), use area code
- Map to IANA timezone identifier

Email: john@company.com.au
TLD: .com.au = Australia
Default: Australia/Sydney

Email: contact@company.de
TLD: .de = Germany
Default: Europe/Berlin

Email: john@company.com
→ Need additional signals (IP, phone, address)

Company headquarters: San Francisco, CA
→ America/Los_Angeles

Company website shows UK address
→ Europe/London

LinkedIn shows "Based in Singapore"
→ Asia/Singapore

Priority:
1. Explicit timezone in data
2. Phone area code
3. Company headquarters
4. Email TLD
5. IP geolocation (least reliable)

Always allow prospect to state preference:

"When's a good time to chat?"
"Mornings work best, I'm in Chicago."
→ Parse and store: America/Chicago, morning preference

Forms:
[Select your timezone ▼]
→ Store as provided

General guidelines (in prospect's local time):

Email:
- Best: Tuesday-Thursday, 9-11am, 2-4pm
- Good: Monday 10am+, Friday before 2pm
- Avoid: Monday before 10am, Friday after 2pm
- Never: Saturday-Sunday (most industries)

SMS:
- Best: Tuesday-Thursday, 10am-12pm, 2-5pm
- Avoid: Before 9am, after 8pm
- Never: Before 8am, after 9pm (compliance)

Phone:
- Best: Tuesday-Thursday, 10-11:30am, 2-4pm
- Good: Right after lunch (1-2pm)
- Avoid: Monday morning, Friday afternoon

B2B Enterprise:
- Email: 8-10am (catch before meetings)
- Calls: 7:30-8:30am (executives available early)

Retail/E-commerce:
- Avoid holiday seasons
- Weekend outreach may work

Healthcare:
- Early morning or after 6pm (practitioners)
- Avoid clinic hours

Startups/Tech:
- Later mornings (10am+)
- Avoid very early (culture tends late)

Finance:
- Very early (6-8am) can work
- Avoid month/quarter end

Learn individual preferences:

Track:
- When do they open emails?
- When do they respond?
- When are calls answered?

Adapt:
- Opens emails at 7am? Send at 6:45am.
- Responds at night? Send EOD.
- Answers calls at lunch? Call at 12:30.

"Send when THEY engage, not when you assume."

{
  "prospect_id": "12345",
  "timezone": {
    "iana": "America/New_York",
    "offset_hours": -5,
    "observes_dst": true,
    "current_offset": -5,
    "detection_method": "phone_area_code",
    "confidence": "high"
  },
  "preferences": {
    "preferred_time": "morning",
    "avoid_days": ["Friday"],
    "stated_by_prospect": true
  },
  "engagement_patterns": {
    "typical_open_hour": 9,
    "typical_response_hour": 14,
    "most_active_day": "Tuesday"
  }
}

def calculate_send_time(prospect, message_type):
    tz = get_prospect_timezone(prospect)

    # Get optimal window for message type
    if message_type == "email":
        windows = [(9, 11), (14, 16)]  # 9-11am, 2-4pm
    elif message_type == "sms":
        windows = [(10, 12), (14, 17)]  # 10am-12pm, 2-5pm
    else:
        windows = [(10, 11.5), (14, 16)]  # Phone

    # Apply individual patterns if available
    if prospect.engagement_patterns:
        preferred_hour = prospect.engagement_patterns.typical_open_hour
        windows = [(preferred_hour, preferred_hour + 1)] + windows

    # Find next available window
    now_local = now_in_timezone(tz)

    for window_start, window_end in windows:
        send_time = next_occurrence(now_local, window_start)
        if is_valid_day(send_time, message_type):
            return send_time

    # Default: next business day, first window
    return next_business_day(now_local, windows[0][0])

def validate_send_time(send_time, prospect, message_type):
    local_time = to_local(send_time, prospect.timezone)

    # Hard limits (legal/compliance)
    if message_type == "sms":
        if local_time.hour < 8 or local_time.hour >= 21:
            return defer_to_next_valid_time(local_time)

    # Soft limits (best practice)
    if local_time.hour < 7 or local_time.hour >= 20:
        return defer_to_next_valid_time(local_time)

    # Weekend handling
    if local_time.weekday() >= 5:  # Saturday or Sunday
        if not prospect.industry_allows_weekend:
            return defer_to_monday(local_time)

    # Holiday handling
    if is_holiday(local_time.date(), prospect.country):
        return defer_to_next_business_day(local_time)

    return send_time

Campaign to 1000 prospects across 10 timezones:

Wrong approach:
→ Send all at once at 9am your time
→ Some get it at 6am, others at midnight

Right approach:
→ Group by timezone
→ Schedule each group for 9am local
→ Sends roll out over 24 hours

Implementation:
1. Group prospects by timezone
2. Calculate send time per group
3. Schedule batches
4. Stagger to avoid delivery issues

For synchronous channels (chat, SMS responses):

If prospect messages at 2am your time:
→ Is it business hours for them?
→ Yes: Respond (bot or human handoff)
→ No: They're night owl, respond anyway

For scheduled messages:
→ Always use their timezone
→ Queue until appropriate window

Sales team across timezones:

Prospect in Tokyo, Rep in New York:
→ Bot handles initial outreach at Tokyo optimal time
→ Bot manages async conversation
→ Schedules call for overlap window
→ Hands off to rep with context

Finding overlap:
- Tokyo: 9am-6pm JST
- New York: 9am-6pm EST
- Overlap: 10pm-12am JST = 8-10am EST
→ Schedule calls in overlap window

Critical: Use IANA timezone IDs, not offsets.

Wrong: Store "UTC-5"
→ Half the year it's UTC-4 (DST)

Right: Store "America/New_York"
→ System handles DST automatically

DST transition handling:
- Check if DST change is upcoming
- Recalculate scheduled sends
- Avoid scheduling exactly at transition hour

Observes DST:
- USA (most), Canada (most)
- Europe
- Australia (some states)
- New Zealand
- Parts of South America

Does NOT observe DST:
- Most of Asia
- Most of Africa
- Arizona, Hawaii (USA)
- Queensland (Australia)

Always check current rules—they change.

SMS/Voice calls:
- Not before 8am local time
- Not after 9pm local time
- Applies to recipient's timezone
- Document timezone determination

Penalties: $500-$1500 per violation

Telemarketing:
- Weekdays: 9am-8pm local
- Saturdays: 9am-5pm local
- Sundays/holidays: Prohibited

SMS similar restrictions apply.

No specific time restrictions, but:
- Consent still required
- Reasonable expectations apply
- 3am emails look suspicious
- Best practice: 8am-8pm local

Regardless of legal requirements:

Safe window:
- 9am-6pm local time
- Monday-Friday
- Avoid local holidays

Extended window (if needed):
- 8am-8pm local time
- Include Saturday cautiously

Never:
- Before 8am local
- After 9pm local
- On local major holidays

Show the prospect you know their timezone:

"Good morning from Sydney!"
(When it's morning there)

"I know it's late in London, so no rush—
whenever you have a moment."
(When sending EOD their time)

"Happy to chat—I'm flexible since I know
we're 12 hours apart. What works for you?"
(Acknowledging the gap)

Always provide escape valve:

"I'll follow up Thursday at 10am your time.
If that doesn't work, let me know what's better."

Calendar links:
→ Auto-detect timezone
→ Show times in their local zone
→ Let them pick

Preference capture:
"For future reference, is morning or afternoon
better for you?"

Test cases:
- Prospect in your timezone (same day)
- Prospect 12 hours ahead (different day)
- Prospect in DST transition
- Prospect in non-DST region
- Unknown timezone prospect

Verify:
- Send time correct in all cases
- No messages outside safe window
- Holiday handling works
- DST transitions handled

Track by timezone:
- Open rates
- Response rates
- Opt-out rates
- Spam complaints

Look for:
- Timezone with poor metrics (wrong detection?)
- Patterns by send hour
- Weekend vs weekday performance

Optimize:
- Adjust send windows per timezone
- Identify timezone-specific preferences
- Fix detection issues

When timezone unknown:

Default strategy:
1. Use company headquarters timezone
2. Or: Assume business hours in most likely region
3. Send at safe middle time (2pm UTC = safe for Americas/Europe)

Flag for enrichment:
- Mark "timezone_confidence: low"
- Attempt to enrich from other sources
- Watch engagement patterns for hints