Payment Funnel Monitor
📖 技能介绍
name: payment-funnel-monitor description: Monitor the payment last mile — initiate → success reconciliation → local method gaps — to recover conversion lost to failed checkouts. Use when payment success rate is below ~95%, specific countries (e.g. Brazil, Germany) underperform vs baseline, support tickets spike for payment timeouts or declines, gateway errors after Shopify/Woo/Stripe updates, or the user asks to fix checkout payment failures. Deliver failure error-code analysis and a prioritized list of local payment methods to add. Also trigger on 3DS/SCA friction, wallet coverage, or "orders drop at pay button." Do NOT use for non-payment checkout fields only (shipping form) with no pay-step failure signal, or pure accounting payout reconciliation with no authorization success rate. compatibility: required: []
Payment funnel monitor
You own pay step → authorization outcome: metrics, error code interpretation, and payment-method completion so fewer sessions die at the last click.
When to lean in
- Success rate < ~95% (define: successful charges / payment attempts or sessions reaching pay — align with merchant).
- Country anomalies (BR, DE, IN, etc.) vs global baseline.
小葱技能7w4.net有完整的技能分类。
- Timeout / decline ticket volume high.
Core workflow
- Reconcile funnel — Initiated payment → submitted → succeeded vs failed vs abandoned (if data exists).
- Error code analysis — Group by gateway code / decline reason; map to action (retry, 3DS, enable local rail, fraud rule).
- Local method strategy — Per underperforming country, recommend wallets, bank transfer, installments, local cards as appropriate.
Gather context
- Platform + gateways (Stripe, Adyen, PayPal, Shopify Payments, etc.).
- Date range; overall and by country success rate.
- Sample error codes or ticket themes.
- Current payment methods enabled per market.
Read references/payment_codes_and_methods.md for code families and country method hints.
Mandatory success outputs (every full run)
1) Failure error code analysis
Structured block — table or bullet groups:
| Error / decline family | Example codes (illustrative) | Likely cause | Action |
|---|---|---|---|
| Insufficient funds | issuer codes | Customer | Messaging; retry |
| 3DS / SCA failed | authentication_required | Bank / UX | 3DS flow, fallback |
| Timeout | gateway timeout | Latency / mobile | Timeout config, alternate method |
| Blocked / fraud | do_not_honor, risk | Rules / issuer | Review rules; local method |
| … | … | … | … |
Use merchant’s actual codes when provided; otherwise use gateway-typical families and label as examples.
2) Recommended local payment methods (checklist)
Per priority country (or globally), output a checklist table:
| Market | Recommended method | Role | Enable via (typical) |
|---|---|---|---|
| Brazil | Pix | Instant, high success | Stripe Pix, PSP |
| Brazil | Boleto (if applicable) | Unbanked | Local PSP |
| Germany | PayPal, SEPA, Klarna | Trust + bank debit | Gateway settings |
| … | … | … | … |
At least three rows when multi-country; tie rows to observed underperformance or generic BR/DE if user named them.
KPI snapshot (include when possible)
| Metric | Value | Target |
|---|---|---|
| Payment success rate | … | ≥95% |
| Success by country (worst 3) | … | vs global |
| Timeout / unknown fail share | … | minimize |
When NOT to use
- Shipping-only abandonment with no payment attempt data.
- Merchant payout settlement with no authorization funnel ask.
Split with other skills
- Abandoned checkout (full funnel) — use when drop is before pay; this skill is pay-step deep.
- Promo traffic stress — use when whole-site CVR drops under load.
🤖 AI 评测
这个 Skill 质量较好，能有效指导支付问题诊断和本地支付方式优化。优点是触发条件明确、分析框架完整、输出要求具体，参考文档实用且覆盖面广。不足是版本早期（v0.1.0），缺少自动化工具支持，主要依赖人工数据收集整理。总体而言适合有技术背景的运营人员使用，可作为支付优化的有效助手。
