charlieviettq/awesome-agent-skill / .claude/skills/algo-price-conjoint
"algo-price-conjoint"
Verified"Run conjoint analysis to measure how product attributes drive consumer preferences and willingness to pay. Use this skill when the user needs to quantify feature value trade-offs, estimate willingness to pay for specific features, or optimize product configuration — even if they say 'which features do customers value most', 'willingness to pay for feature X', or 'product attribute trade-offs'.".
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 4.4 KB |
| examples/sample_scenario.md | markdown | 4.7 KB |
| references/experimental-design.md | markdown | 15.2 KB |
| references/hb-estimation.md | markdown | 12.5 KB |
examples/sample_scenario.md
View on GitHub# Example: SaaS 專案管理工具定價與功能取捨
## Scenario
**公司:** Flowdesk(台灣 B2B SaaS 新創,目標市場:中小企業專案管理)
**PM 的問題:**
> 我們正在規劃 2026 Q3 的新方案改版。現有方案分 Free / Pro ($599/月) / Enterprise。市場研究顯示用戶最在意「AI 自動排程」、「整合數量(Slack/Jira/…)」和「價格」,但我們不知道這三者的相對重要性,也不確定 AI 功能值多少錢。能不能量化一下用戶的 WTP?
**資料:**
- 已完成 CBC 問卷,樣本 N=318(台灣中小企業 IT 採購決策者)
- 設計:3 屬性 × 3 levels,D-optimal 設計,12 選擇題組,每組 3 個方案 + 無購買選項
**屬性與水準:**
| 屬性 | Level 1 | Level 2 | Level 3 |
|------|---------|---------|---------|
| AI 自動排程 | 無 | 基礎(建議) | 全自動 |
| 整合數量 | 5 個 | 15 個 | 無限 |
| 月費(/用戶) | $299 | $599 | $999 |
---
## Analysis
### Phase 1: 輸入驗證
- **屬性獨立性:** AI 排程與整合數量互不依賴 ✓
- **水準現實性:** $999 是市場上游方案實際售價(Asana Business),不是假設 ✓
- **樣本:** N=318 > 200 最低門檻 ✓
- **設計:** D-optimal 12 task × 3 alternatives,每屬性在每位置出現頻率均衡 ✓
**Gate passed.**
---
### Phase 2: 模型估計(Hierarchical Bayes)
使用 HB-MNL 估計個體層級 part-worth utilities(先驗:normal hierarchical prior,MCMC 10,000 次迭代,burn-in 2,000)。
**Part-worth utilities(群體平均):**
| 屬性 | 水準 | Part-worth |
|------|------|-----------|
| AI 排程 | 無 | −1.42 |
| AI 排程 | 基礎(建議) | +0.61 |
| AI 排程 | 全自動 | +1.89 |
| 整合數量 | 5 個 | −0.83 |
| 整合數量 | 15 個 | +0.44 |
| 整合數量 | 無限 | +0.97 |
| 月費 | $299 | +2.31 |
| 月費 | $599 | +0.18 |
| 月費 | $999 | −2.14 |
**屬性重要性計算:**
- AI 排程 range = 1.89 − (−1.42) = **3.31**
- 整合數量 range = 0.97 − (−0.83) = **1.80**
- 月費 range = 2.31 − (−2.14) = **4.45**
- 總和 = 9.56
| 屬性 | 重要性 |
|------|--------|
| 月費 | 46.5% |
| AI 排程 | 34.6% |
| 整合數量 | 18.8% |
**WTP 推導(utility-to-price conversion):**
price coefficient(每 $100)= (2.31 − (−2.14)) / (($999 − $299) / $100) = 4.45 / 7 = **0.636 utils / $100**
- WTP for AI 排程(無 → 全自動)= 3.31 / 0.636 × $100 = **$520/用戶/月**
- WTP for AI 排程(無 → 基礎)= 2.03 / 0.636 × $100 = **$319/用戶/月**
- WTP for 整合(5 → 無限)= 1.80 / 0.636 × $100 = **$283/用戶/月**
---
### Phase 3: 驗證
**Holdout task(保留 2 題未納入估計):**
- Holdout hit rate = **71.4%**(> 60% 門檻 ✓)
- 符號檢查:價格越高效用越低(−2.14 → −0.18 → +2.31,即 $999 < $599 < $299)✓
- AI 全自動 > 基礎 > 無 ✓
**Gate passed.**
---
### 潛在細分(Latent Class,2-segment 解)
| 區隔 | 比例 | 特徵 | AI 排程重要性 | 價格重要性 |
|------|------|------|--------------|-----------|
| 功能導向(Segment A) | 58% | 軟體/科技業 IT | 41% | 35% |
| 價格敏感(Segment B) | 42% | 傳統產業採購 | 22% | 58% |
> **關鍵發現:** Segment A 的 WTP for AI 全自動 ≈ $680,Segment B 僅 ≈ $310。
---
## Result
```json
{
"attribute_importance": [
{"attribute": "price", "importance_pct": 46.5},
{"attribute": "ai_scheduling", "importance_pct": 34.6},
{"attribute": "integrations", "importance_pct": 18.8}
],
"part_worths": {
"ai_scheduling": {"none": -1.42, "basic": 0.61, "full_auto": 1.89},
"integrations": {"5": -0.83, "15": 0.44, "unlimited": 0.97},
"price": {"$299": 2.31, "$599": 0.18, "$999": -2.14}
},
"wtp": {
"ai_scheduling_none_to_full_auto": 520,
"ai_scheduling_none_to_basic": 319,
"integrations_5_to_unlimited": 283
},
"metadata": {
"respondents": 318,
"model": "hierarchical_bayes",
"holdout_hit_rate": 0.714,
"segments": 2
}
}
```
**給 PM 的三點建議:**
1. **AI 全自動排程是最大差異化點**,WTP $520 遠超功能開發成本預估。建議做為新 Pro+ 方案的核心賣點,定價 $799/月仍在 Segment A(58% 市場)的接受區間內。
2. **整合數量效益邊際遞減**:從 5→15 個整合的 WTP ≈ $126,但 15→無限的增量 WTP 僅 ≈ $83。維持「Pro 方案 15 個整合」即可,無限整合留給 Enterprise,不需要在 Pro 層放滿。
3. **勿外插到 $299 以下或 $999 以上**:本研究設計範圍 $299–$999,在此範圍外的 WTP 估計不可靠——若考慮免費方案轉換率,需另設計 Free vs. Paid 的獨立 CBC 研究。
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis. | Stars: 472k | Updated | Claude CodeCodex | Verified | Install |
| diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness. | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |
| executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is available | Stars: 323k | Updated | Claude CodeCodex | Verified | Install |