中
✅
Scan Success!
View 客户 journey timeline profile
Redirecting in 3 seconds
High-value Behavior
View Triggered Tasks
Behavior
Day 7 High-Value Behavior: Requested callback from success team
Engagement Level
High
Repeated views indicate strong intent
Pool Status
C
Currently in C Pool
AI Score Growth Animation (0 → 7)
B Pool Threshold: 60
Day 1
0
Day 3
0
Day 7 (Actual)
7
Upgrade Status
Below Threshold
Customer is currently in C Pool.
📊 Real-time Score Statistics
Today's Total Score
0
All behaviors combined
Event Type Score Details
No behavior records today
📋 Detailed Scoring Rules
▼

Base Score Rules

View Document
View product documentation
+10 分
Watch Video
Watch video (min 5s)
+30 分
Repeat Watch
Repeat watch video
+30 分
Contact Support
Contact support/request callback
+5 分
Translation Feature
Click translation feature card
+15 分
Culture Insight
Click culture insight feature card
+15 分
Health Monitoring
Click health monitoring feature card
+10 分
Remote Assistance
Click remote assistance feature card
+15 分
Industry Selection
Select industry solution
+8 分
Meeting Scenario
Answer meeting scenario (wrong answer: 50%)
+12 分
Card Etiquette
Answer card etiquette scenario (wrong answer: 50%)
+12 分
Survey Complete
Complete post-video survey
+20 分
Expo Form
Submit expo H5 registration form
+25 分

Same-day Decay Rules

第1次
100%
(30 pts)
第2次
67%
(20 pts)
第3次
33%
(10 pts)
第4次
17%
(5 pts)
第5次+
0%
(0 pts)
Note: Video watching behaviors (Watch Video, Repeat Watch) decay according to the above ratios when triggered multiple times in one day. Other event types use different decay rules. Contact Support is immune to decay.

Anti-flood Mechanism

Time Window: 5 分钟
Rule: If the same event type is triggered again within 5 minutes (and it's the first time today), the score will be set to 0. Decayed scores are not affected.
Exception: CONTACT_SUPPORT is immune to anti-flood mechanism.
AI客户培育引擎解读
- **High-value signal, low proof of intent.** A callback request from the success team is a direct, human-initiated buying signal — it shortens the sales cycle by moving from anonymous browsing to a named conversation. / 主动要求成功团队回电是明确的购买信号,把匿名浏览变成有名有姓的对话,直接压缩销售周期。 - **Score jump is real but insufficient.** Score moved 0 → 7 (from zero baseline), a strong relative lift, yet 7 vs. the 60 threshold means the AI sees intent without qualification depth — no budget, scope, or decision-role data yet. / 分数从 0 跳到 7,相对增幅显著,但距 60 分门槛仍远:AI 判断有意图、缺资质信息(预算、规模、决策角色)。 - **Why it stays in C Pool.** Pool upgrades are threshold-driven, not sentiment-driven; keeping it in C prevents premature sales load and preserves routing efficiency until firmographic and need data arrive. / 分层靠阈值而非感觉;留在 C 池避免过早占用销售资源,等企业与需求信息补齐后再升级。 - **Triggered actions now.** (a) Success team callback within SLA (e.g., 4 business hours); (b) capture qualification fields during the call (units, timeline, training audience); (c) auto-log call notes to feed the scoring model; (d) schedule a Day 10 re-score checkpoint. / 立即触发:成功团队按 SLA 回电;通话中补齐资质字段(数量、时间表、受训对象);通话记录回写评分模型;设置第 10 天复评节点。 - **Path to the 100-unit inquiry.** A qualified callback typically lifts score past 60 → C→B upgrade → proposal and sales follow-up tasks fire → the 100-unit distribution deal gets a structured proposal and owner. Without the callback data, that conversion stays unforecastable. / 有效回电通常能把分数推过 60,触发 C→B 升级、方案与销售跟进任务,从而把 100 套分销询盘变成有方案、有负责人的商机;缺回电数据则无法预测转化。