drip-pacing-intelligence

ドリップペーシングインテリジェンス

営業オートメーションにおけるインテリジェントなシーケンスのタイミング設定を専門としています。見込み客のエンゲージメントに応じてアプローチのペースを動的に調整するシステムの開発を支援します。

ペーシングが重要な理由

固定ペーシングの問題

Fixed schedule:
Day 1 → Day 3 → Day 7 → Day 14

What happens:
- Hot prospect: Cooled off waiting 7 days
- Cold prospect: Annoyed by Day 3 message
- Busy prospect: Missed all messages during travel

Result: Lower response rates, higher opt-outs

インテリジェントなペーシングのメリット

Dynamic adjustment:
- Engaged? → Accelerate
- Quiet? → Decelerate
- Signals? → Respond immediately

Result: Right message, right time, right frequency

エンゲージメントシグナル

ポジティブなシグナル（加速）

High intent:
- Email opened (multiple times)
- Links clicked
- Attachment downloaded
- Website visited
- Content engaged with
- Partial form filled

Action: Shorten intervals, add touches

中立的なシグナル（維持）

Normal behavior:
- Email opened once
- No response yet
- Profile viewed (LinkedIn)
- Website bounce

Action: Stay on current pace

ネガティブなシグナル（減速）

Low engagement:
- No opens
- Unsubscribes from marketing
- Marked as spam
- Auto-reply received

Action: Lengthen intervals, reduce touches

ペーシングアルゴリズム

エンゲージメントスコアに基づくペーシング

Calculate engagement score (0-100):
+10: Email opened
+20: Link clicked
+30: Website visit
+40: Content downloaded
+50: Reply received
-10: Email ignored (after 48h)
-20: Unsubscribe
-50: Spam complaint

Pacing rules:
Score > 70: Next touch in 24h
Score 40-70: Next touch in 48-72h
Score 20-40: Next touch in 5-7 days
Score < 20: Pause sequence, try different approach

適応型の間隔計算

def calculate_next_touch(prospect):
    base_interval = prospect.sequence.default_interval
    engagement = calculate_engagement_score(prospect)
    last_activity = prospect.last_engagement_at

    # Engagement multiplier
    if engagement > 70:
        multiplier = 0.5  # Half the time
    elif engagement > 40:
        multiplier = 1.0  # Normal pace
    elif engagement > 20:
        multiplier = 1.5  # Slow down
    else:
        multiplier = 2.5  # Significantly slow

    # Recency adjustment
    hours_since_activity = hours_since(last_activity)
    if hours_since_activity < 4:
        # Very recent activity—strike while hot
        return timedelta(hours=2)

    interval = base_interval * multiplier
    return interval

イベントトリガーによる加速

Immediate response triggers:
- Website pricing page visit → Trigger within 1 hour
- Demo video watched >50% → Follow up same day
- Competitor comparison viewed → Priority outreach
- Multiple page views in session → Real-time engagement

"I noticed you were just checking out our pricing—
any questions I can answer?"

シーケンス調整パターン

加速パターン

Standard sequence:
Day 1 → Day 4 → Day 8 → Day 14

Engagement detected (Day 2 email opened, link clicked):
Day 1 → [Engagement] → Day 2 (accelerated) → Day 4 → Day 8

New sequence:
Day 1 → Day 2 → Day 5 → Day 10

Message adjustment:
"I saw you checked out [content]. Wanted to follow up
while it's fresh—did that answer your questions?"

減速パターン

Standard sequence:
Day 1 → Day 4 → Day 8 → Day 14

No engagement detected:
Day 1 → Day 4 (no open) → Day 10 (extended) → Day 20

Message adjustment:
"I've reached out a couple times—I know inboxes get
crowded. If the timing isn't right, no worries. But
if [problem you solve] is on your radar, I'd love to
help when you're ready."

急増パターン

High engagement burst detected:
- 3 emails opened in 1 hour
- Website visit
- LinkedIn profile view

Trigger immediate outreach:
"Hey [Name], looks like you're doing some research—
perfect timing. Would a quick call be helpful, or
do you prefer I send some info?"

マルチチャネルのペーシング

チャネルの連携

Don't pile on. Coordinate across channels:

Day 1: Email
Day 3: Email (if opened Day 1)
Day 3: LinkedIn (if no open on Day 1)
Day 6: SMS (if engagement on either)
Day 8: Call (if high engagement)

Never:
- Email + SMS + Call same day
- Multiple touches in same channel < 48h apart

チャネルごとのペーシング

Email:
- Minimum 48h between sends
- Max 3 emails per week
- Extend if no opens

SMS:
- Minimum 3 days between sends
- Max 2 per week
- Stop if no response after 2

Phone:
- Minimum 5 days between attempts
- Max 3 per sequence
- Voicemail counts as attempt

タイムゾーンへの対応

最適な送信時刻

Adjust for recipient timezone:
- Morning: 8-10am local
- Afternoon: 2-4pm local
- Never: Outside 8am-8pm local

If timezone unknown:
- Infer from area code
- Infer from email domain
- Default to safe window

週末・祝日のペーシング

Pause logic:
- No outreach Saturday-Sunday
- Check for holidays
- Resume Monday morning

Adjustment:
If Day 4 falls on Saturday:
→ Move to Monday (Day 6)
→ Adjust subsequent touches proportionally

シーケンスの変更

動的なシーケンス選択

Based on engagement, switch sequences:

Low engagement (Score < 30):
→ Switch to "Break-up" sequence
→ Fewer touches, more spaced out
→ Different messaging approach

High engagement (Score > 70):
→ Switch to "Fast-track" sequence
→ More frequent touches
→ More direct CTAs
→ Add phone touches

タッチタイプの変更

Based on engagement patterns:

Opens but no clicks:
→ Make next email shorter
→ Stronger CTA
→ Single focus

Clicks but no response:
→ Add different channel
→ Address potential objection
→ Offer easier next step

実装

状態管理

{
  "prospect_id": "12345",
  "sequence_id": "enterprise_outbound",
  "current_step": 3,
  "engagement_score": 65,
  "pacing_status": "accelerated",
  "next_touch_at": "2024-01-16T10:00:00Z",
  "last_engagement": {
    "type": "email_click",
    "timestamp": "2024-01-15T14:30:00Z",
    "content": "pricing_link"
  },
  "pacing_history": [
    {"step": 1, "interval_days": 0, "actual_days": 0},
    {"step": 2, "interval_days": 3, "actual_days": 2},
    {"step": 3, "interval_days": 3, "actual_days": 2}
  ]
}

ペーシングルールエンジン

rules = [
    {
        "condition": "email_opened_count >= 3 AND hours_since_last_open < 24",
        "action": "accelerate",
        "modifier": 0.5
    },
    {
        "condition": "no_opens AND step >= 3",
        "action": "decelerate",
        "modifier": 2.0
    },
    {
        "condition": "website_visit AND page == 'pricing'",
        "action": "immediate_touch",
        "delay_hours": 2
    },
    {
        "condition": "out_of_office_detected",
        "action": "pause",
        "resume_date": "ooo_return_date + 1"
    }
]

指標と最適化

ペーシングの有効性

Measure:
- Response rate by pacing tier
- Opt-out rate by pacing tier
- Time to conversion by pacing
- Engagement score trends

Compare:
- Static vs dynamic pacing
- Different pacing algorithms
- A/B test interval changes

継続的な学習

Track outcomes:
- Which adjustments led to responses?
- What pacing converts best?
- When do prospects disengage?

Feed back:
- Update pacing rules
- Refine engagement scoring
- Improve timing predictions