← 返回 Skills 市场
leooooooow

Loyalty Designer

作者 LeroyCreates · GitHub ↗ · v1.1.0 · MIT-0
cross-platform ✓ 安全检测通过
171
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install loyalty-designer
功能描述
Design points-based, tiered, or referral loyalty programs with reward structures calibrated to your margin and customer purchase frequency.
使用说明 (SKILL.md)

Loyalty Designer

Customer loyalty programs are margin investments — done right they increase purchase frequency and LTV; done wrong they're expensive discount machines that train customers to wait for rewards. Loyalty Designer helps you build a complete loyalty program architecture from scratch: points structures, tier thresholds, reward catalog, referral mechanics, and the financial model that determines whether the program actually improves contribution margin.


Quick Reference

Decision Strong Acceptable Weak
Program type for low-AOV repeat buyers Points-based (frequent earn/burn) Tiered with low entry threshold Referral-only (misses existing behavior)
Program type for high-AOV infrequent buyers Tiered VIP (status + perks) Cash back Points (too slow to build engagement)
Points earn rate 1-5% of spend as points value 6-8% of spend > 10% (margin-destroying)
Tier threshold design Based on actual purchase frequency data Estimated from AOV × target visits Arbitrary round numbers
Reward breakage estimate 20-35% expected (industry norm) 10-20% conservative 0% (always dangerous assumption)
Referral reward structure Dual-sided (referrer + referee) Referrer only No referral mechanic

Solves

This skill addresses these specific problems:

  1. High one-time buyer rate — Most customers buy once and never return; no structured incentive exists to reward repeat behavior.
  2. Unsustainable discount programs — Store-wide discount codes replace loyalty structures, training customers to buy only on sale and crushing margin.
  3. No differentiation between high-LTV and low-LTV customers — Everyone receives the same treatment regardless of how much they've spent.
  4. Referral programs that don't launch — Referral mechanics designed as afterthoughts with no realistic incentive calculation or tracking plan.
  5. Loyalty programs that lose money — Programs launched without a break-even analysis; earn rates set too generously relative to product margin.
  6. Low engagement after enrollment — Customers sign up for loyalty programs but never redeem, often because reward thresholds are too high or communication is absent.
  7. Fragmented program mechanics — Points, tiers, and referrals designed as separate features rather than an integrated system reinforcing each other.

Workflow

Step 1 — Define program type based on business model

Match the program structure to purchase behavior:

Points-based programs work best when:

  • AOV is under $100 and customers can realistically purchase 3+ times per year
  • Product category has natural repurchase cycles (consumables, apparel basics, pet supplies)
  • You want to reward every transaction and build a habit of earning

Tiered programs work best when:

  • You have a meaningful range of customer spend levels ($200 to $2,000+/year)
  • You want to create aspiration and status differentiation
  • High-tier customers should receive service-level perks (early access, dedicated support) not just discounts

Referral programs work best when:

  • Customer acquisition cost (CAC) is high relative to LTV
  • Your product has strong word-of-mouth potential (new problem solved, visible product, gift-able item)
  • Referral mechanic supplements an existing retention program (not a replacement for it)

Many effective programs combine types: a points foundation with tier status overlaid and a referral accelerator.

Step 2 — Establish the financial model

Before designing any earn/burn rates, calculate what the program can afford:

Max Program Cost % = Gross Margin % − Target Contribution Margin %

Example: 45% gross margin, target 38% contribution after loyalty costs
Max Program Cost = 7% of revenue

Within that 7%, allocate:

  • Points redemption cost: 3-4% (accounting for 25-30% breakage)
  • Tier reward costs: 1-2%
  • Referral incentive cost: 1-2%

Breakage (points earned but never redeemed) is critical to model accurately. Industry average is 25-35%. Program designs that assume 0% breakage consistently lose money.

Step 3 — Design the points structure

Earn rate:

Points earned per dollar = Target redemption value / (100 − breakage %)

If 100 points = $1 reward and 30% breakage:
Customer earns $0.70 in real value per 100 points (you owe $0.70 per 100 pts earned)
Effective cost per dollar spent = earn rate × $0.70

Standard earn rates by reward value:

  • 1 point per $1 spent = $0.01 per point if 100 pts = $1 redemption
  • Common: 5 points per $1 with 500 pts minimum redemption = ~1% value to customer

Minimum redemption threshold: Set high enough to encourage repeat purchases before redeeming. Target: redemption value equal to 1.5-2× AOV spend required to earn it.

Point expiry: Add 12-18 month expiry to manage liability and create urgency. Always notify customers 30 days before expiry.

Step 4 — Design tier thresholds and benefits

Tier thresholds should be based on actual purchase data:

Tier 1 threshold = 12-month spend at or above 40th percentile of active customers
Tier 2 threshold = 12-month spend at or above 75th percentile
Tier 3 threshold = 12-month spend at or above 90th percentile

If no data is available, use AOV × estimated annual purchases:

  • Bronze: 1-2 purchases per year
  • Silver: 3-5 purchases per year
  • Gold: 6+ purchases per year or total spend > $X

Tier benefits structure:

Tier Points Multiplier Discount Service Perk Access Perk
Bronze None needed Standard Standard
Silver 1.5× 5-10% on select Priority support Early sale access
Gold 10-15% on select Dedicated line Pre-launch access

Benefits should include at least one non-discounted perk per tier to avoid pure discount training.

Step 5 — Design the referral mechanic

Dual-sided referral (always preferred):

  • Referrer receives: reward triggered when referee makes first purchase
  • Referee receives: discount or bonus on first order

Referral economics:

Max referral reward = Gross Margin on first referee order − New customer CAC avoided

If CAC is $40 and gross margin on first order is $25 → referral reward should not exceed $25 (to avoid spending more than CAC already costs).

Referral tracking requirements:

  • Unique referral codes or links per customer
  • Attribution window: typically 30 days from link share to first purchase
  • Fraud protection: limit referrals per account, restrict same-address referrals

Step 6 — Build the reward catalog

Reward options by cost effectiveness:

Reward Type Margin Impact Customer Perceived Value Recommended
Discount on future order High (direct margin cost) Medium Limit to % of catalog
Free product (own product) Medium (COGS only) High Strong option
Free shipping threshold removal Low (variable) High for frequent buyers Yes
Early access / experiences Very low High for top tier Yes for Gold tier
Third-party gift cards Fixed cost Medium Use sparingly

Step 7 — Define the communication and engagement calendar

Loyalty programs fail without ongoing engagement communication:

  • Enrollment confirmation: Points balance, how to earn, redemption instructions
  • Points milestone emails: Triggered at 25%, 50%, 75%, and 100% of redemption threshold
  • Expiry warnings: 30 days, 7 days before point expiry
  • Tier upgrade notification: Celebrate achievement, show next tier benefits
  • Tier downgrade warning: 30 days before end of qualifying period
  • Referral nudge: After 2nd purchase, remind customer of referral program

Worked Examples

Example 1 — Skincare Brand (High-Repeat, Low-AOV)

Inputs:

  • Category: Skincare (moisturizers, serums, cleansers)
  • AOV: $45
  • Purchase frequency: ~4x/year for active customers
  • Gross margin: 62%
  • Current one-time buyer rate: 68%
  • Goal: Increase repeat purchase rate to 45%

Program design:

Points structure:

  • 5 points per $1 spent
  • 500 points = $5 reward (1% effective return to customer)
  • Minimum redemption: 500 points ($100 spend required → 2.2 orders to first reward)
  • 18-month expiry

Tiers:

  • Bronze (default): 0–$149/year spend
  • Silver: $150–$299/year (3–4 orders) → 1.5× points, birthday gift
  • Gold: $300+/year (6+ orders) → 2× points, free shipping always, early access

Referral:

  • Referrer: 250 bonus points (~$2.50 value) on referee's first order
  • Referee: $5 off first order
  • Max referral reward: $7.50 total (vs. $22 blended CAC)

Financial model:

  • Points earn cost: 1% × (1 − 30% breakage) = 0.7% of revenue
  • Tier perks: 1.2% of revenue estimated
  • Referral cost: 0.8% of revenue at projected referral volume
  • Total loyalty cost: 2.7% of revenue (well within 7% max)

Example 2 — Home Goods Brand (Low-Frequency, High-AOV)

Inputs:

  • Category: Furniture and home décor
  • AOV: $320
  • Purchase frequency: 1-2×/year max for most customers
  • Gross margin: 48%
  • Goal: Increase referrals and encourage upsell categories

Program design:

Tier structure (points too slow for annual buyers):

  • Member: 0–$499/year → Free returns, style consultation access
  • Insider: $500–$999/year → 5% back on next purchase, priority delivery, early sale
  • Elite: $1,000+/year → Personal stylist, white-glove delivery, exclusive catalog access

No standard points: Replace with "purchase credit" — 3% of every order credited to account, redeemable on orders $200+. This avoids the "points feel cheap" problem for luxury positioning.

Referral:

  • Referrer: $30 account credit when referee spends $200+
  • Referee: 10% off first order
  • Financial check: Referee first order $320 → 10% discount = $32 cost. Plus $30 referral credit = $62 total. Gross margin on $320 = $153. Net margin on referred order: $153 - $62 = $91. Positive even on first order.

Program positioning: Not called a "loyalty program" — positioned as a "Home Collective membership" to match premium brand positioning.


Common Mistakes

  1. Setting earn rates before calculating margin math — Many programs launch at 2-5% customer value before checking whether that's sustainable at scale.

  2. Zero breakage assumption — Assuming all points will be redeemed. Industry data shows 25-35% of points are never redeemed. Building this into financials reduces required earn rates.

  3. Tier thresholds too easy to reach — If 70% of customers immediately qualify for Silver, Silver has no aspirational value and you're giving Silver perks to everyone.

  4. Discount-only reward catalogs — Training every customer to seek discounts. Mix discount rewards with experience and access rewards to protect margin and increase perceived program value.

  5. No engagement communication plan — Launching a program without points milestone emails means most enrolled customers forget they're in it.

  6. Single-sided referral programs — Referral programs that only reward the referrer (and not the new customer) consistently underperform because the referee has no incentive to act.

  7. Points expiry that's too short — 6-month expiry feels punitive and drives disengagement. 12-18 months is standard; expire too fast and customers opt out entirely.

  8. Program design doesn't match brand positioning — A luxury brand calling it a "points program" with a leaderboard cheapens perception. Program design must match brand voice.

  9. No fraud prevention — Referral programs without same-address restrictions or account limits quickly attract abuse from customers self-referring or creating duplicate accounts.

  10. Launching without a sunset plan — If the program doesn't achieve retention goals after 12 months, you need a way to end or restructure it without alienating enrolled customers.


Resources

  • references/output-template.md — Full loyalty program design output format
  • references/program-economics-guide.md — Financial modeling for points, tiers, and referrals
  • references/tier-design-benchmarks.md — Industry benchmarks for thresholds and benefit structures
  • assets/loyalty-design-checklist.md — Pre-launch and quarterly program health checklist
安全使用建议
This skill is a documentation/template package for designing loyalty programs and appears internally consistent. Things to consider before installing: (1) provenance — the source and owner are unknown, so confirm you trust the provider before using company-sensitive data; (2) do not paste real customer PII, credentials, or raw financial exports into the skill's prompts or outputs unless you control where that data goes; (3) validate the financial model outputs against your actual margin and transaction data before acting on recommendations; and (4) if you plan to integrate any of these designs with production systems (points ledger, referral attribution), audit the actual implementation code or vendor you will use — the guidance here is conceptual and must be implemented securely.
功能分析
Type: OpenClaw Skill Name: loyalty-designer Version: 1.1.0 The 'Loyalty Designer' skill bundle is a purely instructional and template-based resource for designing business loyalty programs. It contains no executable code, shell commands, or network activity. The content across all files, including SKILL.md and the reference guides, is focused entirely on financial modeling, tier structures, and marketing strategy without any indicators of malicious intent or prompt injection risks.
能力评估
Purpose & Capability
The name/description (designing points, tiers, referrals) matches the provided SKILL.md and reference files. There are no unrelated env vars, binaries, or install steps requested — everything is documentation and templates appropriate for a design/consulting skill.
Instruction Scope
SKILL.md and the supporting files contain only procedural guidance, formulas, templates, and checklists for designing loyalty programs. There are no commands, system file reads, external endpoints, or instructions to access credentials. The runtime instructions stay within the stated design scope.
Install Mechanism
No install spec or code files are present; this is instruction-only. That minimizes disk/writing/execution risk and is proportionate for a documentation/template skill.
Credentials
The skill does not request environment variables, credentials, or config paths. The materials do not require secrets or external service access; requested resources are proportionate to the stated purpose.
Persistence & Privilege
always is false and the skill does not request elevated or persistent system privileges. It does not modify other skills' configs or require permanent presence.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install loyalty-designer
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /loyalty-designer 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.1.0
v1.1.0: Added tier design benchmarks reference, program economics guide, and output template. Expanded checklist with quarterly health review and red flags section.
v1.0.0
Initial release.
元数据
Slug loyalty-designer
版本 1.1.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 2
常见问题

Loyalty Designer 是什么?

Design points-based, tiered, or referral loyalty programs with reward structures calibrated to your margin and customer purchase frequency. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 171 次。

如何安装 Loyalty Designer?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install loyalty-designer」即可一键安装,无需额外配置。

Loyalty Designer 是免费的吗?

是的,Loyalty Designer 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Loyalty Designer 支持哪些平台?

Loyalty Designer 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Loyalty Designer?

由 LeroyCreates(@leooooooow)开发并维护,当前版本 v1.1.0。

💬 留言讨论