Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →基于客户资产规模、贡献度、活跃度、产品持有等多维度信息，识别客户价值层级，输出分层结论与经营建议，支撑差异化客户经营策略。
.claude/skills/aifinlab-customer-segmentation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-08 | ✗→✓ | ▲ Improved | 58% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 103% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 13% | 0% |
银行零售客户数量庞大，资源有限，需要对客户进行价值分层以实现差异化经营。传统分层仅依赖AUM单一维度，忽略了客户贡献度、成长潜力、活跃度和忠诚度等重要因素，导致营销资源错配、高潜力客户流失。需要多维度综合评估客户价值层级，支撑精细化客户经营。
| 输入项 | 说明 | 是否必填 | |--------|------|----------| | 客户列表 | 待分层的客户编号集合 | 必填 | | AUM及资产结构 | 客户总资产、存款、理财、基金等构成 | 必填 | | 贡献度指标 | 中间业务收入、利息贡献、综合贡献 | 选填 | | 产品持有信息 | 持有产品种类和数量 | 选填 | | 交易活跃度 | 交易频次、登录频次、渠道使用情况 | 选填 | | 客户关系时长 | 开户时间、持续在册年数 | 选填 | | 成长趋势 | 近3/6/12个月AUM变化趋势 | 选填 | | 风险标签 | 逾期、投诉、异常交易等标记 | 选填 |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | fail→pass | 20,479 | 25,615 | +25% | 1 | 1 | 0% | 3,105 | 4,895 | +58% | 0 | 0 | — |
case-01 | fail→pass | 24,283 | 32,467 | +34% | 1 | 1 | 0% | 3,636 | 4,876 | +34% | 0 | 0 | — |
case-02 | fail→pass | 20,971 | 30,946 | +48% | 1 | 1 | 0% | 3,190 | 6,479 | +103% | 0 | 0 | — |
case-03 | fail→pass | 25,037 | 28,904 | +15% | 1 | 1 | 0% | 3,995 | 5,528 | +38% | 0 | 0 | — |
case-04 | fail→pass | 20,902 | 17,142 | -18% | 1 | 1 | 0% | 3,166 | 3,593 | +13% | 0 | 0 | — |
case-05 | fail→fail | 16,874 | 12,118 | -28% | 1 | 1 | 0% | 2,616 | 2,805 | +7% | 0 | 0 | — |
case-06 | fail→pass | 12,039 | 16,174 | +34% | 1 | 1 | 0% | 1,844 | 3,317 | +80% | 0 | 0 | — |
case-07 | fail→pass | 13,880 | 11,660 | -16% | 1 | 1 | 0% | 1,903 | 2,817 | +48% | 0 | 0 | — |
case-09 | pass→pass | 19,648 | 13,478 | -31% | 1 | 1 | 0% | 2,851 | 3,091 | +8% | 0 | 0 | — |
case-10 | pass→pass | 19,241 | 18,850 | -2% | 1 | 1 | 0% | 2,766 | 3,875 | +40% | 0 | 0 | — |
case-11 | pass→pass | 14,747 | 12,773 | -13% | 1 | 1 | 0% | 2,234 | 3,093 | +38% | 0 | 0 | — |
case-12 | pass→pass | 16,326 | 13,780 | -16% | 1 | 1 | 0% | 2,406 | 3,203 | +33% | 0 | 0 | — |
case-13 | pass→pass | 12,896 | 13,301 | +3% | 1 | 1 | 0% | 1,857 | 3,013 | +62% | 0 | 0 | — |
case-14 | fail→pass | 22,650 | 34,323 | +52% | 1 | 1 | 0% | 3,693 | 7,164 | +94% | 0 | 0 | — |
case-15 | pass→pass | 16,350 | 13,810 | -16% | 1 | 1 | 0% | 2,364 | 3,024 | +28% | 0 | 0 | — |
case-16 | pass→pass | 16,184 | 12,223 | -24% | 1 | 1 | 0% | 2,399 | 2,861 | +19% | 0 | 0 | — |
case-17 | pass→pass | 18,591 | 17,477 | -6% | 1 | 1 | 0% | 2,739 | 3,555 | +30% | 0 | 0 | — |
case-18 | fail→pass | 15,874 | 15,338 | -3% | 1 | 1 | 0% | 2,675 | 3,469 | +30% | 0 | 0 | — |
case-19 | fail→pass | 15,845 | 4,095 | -74% | 1 | 1 | 0% | 2,485 | 1,765 | -29% | 0 | 0 | — |
case-20 | fail→pass | 15,992 | 8,126 | -49% | 1 | 1 | 0% | 2,451 | 2,306 | -6% | 0 | 0 | — |
case-21 | fail→pass | 19,468 | 8,957 | -54% | 1 | 1 | 0% | 2,939 | 2,329 | -21% | 0 | 0 | — |
case-22 | fail→pass | 13,530 | 9,975 | -26% | 1 | 1 | 0% | 1,807 | 2,551 | +41% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +59 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.