Loyalty Tier
by
Tier design for ecommerce teams: set thresholds where they change behavior and rewards that cost less than the behavior is worth.
Loyalty tiers usually get set at round numbers with rewards chosen by what competitors offer, which produces a program that pays your best customers for what they were doing anyway. The two decisions that matter are where the thresholds sit and what sits behind each one. A threshold just above where customers naturally cluster pulls people up; one below that cluster hands a reward to everyone already there. A reward that costs more than the behavior change it buys is a discount with extra steps. This skill sets thresholds where they change behavior and rewards that cost less than they earn.
What It Does
- Reads where your customers actually cluster in annual spend.
- Places thresholds where a reachable stretch exists rather than at round numbers.
- Prices each reward against the incremental behavior it is expected to buy.
- Estimates how many customers would move up versus how many already qualify.
- Prefers a non-discount reward wherever it performs comparably.
- Recommends the thresholds and the reward behind each, with the net.
Use Cases
- Set a threshold that pulls customers up rather than rewarding the arrived.
- Replace a discount tier with a benefit that costs less and works as well.
- See how many customers are already above each proposed threshold.
- Price a loyalty program before launching it.
Example Session
Design our loyalty tiers. What thresholds and what rewards?
Your customers cluster at $180 and $440. Put the thresholds just above both.
- Tier | Threshold | Already above | Would stretch | Reward | Net per year
- Silver | $200 | 4,100 | 2,800 | Free shipping, always | +$61,000
- Gold | $475 | 900 | 1,100 | Early access plus free returns | +$88,000
- Platinum | $1,200 | 140 | 90 | Named contact, first allocation | +$34,000
- Rejected: 10% off at $150 | $150 | 6,900 | 300 | 10% off | -$112,000
The rejected tier is the one most programs launch: a round threshold below where people already sit, so 6,900 customers get 10% off for nothing and 300 stretch. Setting Silver at $200 instead means 2,800 customers have a reachable reason to add an item, and free shipping costs you less than the discount would on every one of them.
Why This Beats Prompting It Yourself
A Round Threshold vs Where Customers Actually Cluster: The Hidden Risk
A generic prompt proposes tiers at $100, $500 and $1,000, because those are the numbers that sound like tiers. If your customers cluster at $180, a $100 threshold rewards nearly everyone for nothing. The distribution decides where a threshold works, and it is never at a round number.
When You Still Need The Reward To Cost Less Than It Buys
Every tier has two numbers: what it costs on the customers already above it, and what it earns from the ones who stretch. A 10% discount tier fails that test almost always, because the customers already qualifying vastly outnumber the ones who move. Free shipping and early access pass it, which is why they appear here and the discount does not.
Known Limitations
Stretch estimates come from how your customers have responded to previous thresholds, and a store that has never run tiers gets a wider band the output labels. Non-discount rewards depend on you actually having something worth early access to. It recommends the design and does not configure a loyalty platform.
Install It
- Download the ZIP.
It is free and there is no account to create. - Unzip it into your agent's skills folder.
Claude Code reads ~/.claude/skills/, which is hidden by default: the command in the folder block opens it. Other agents scan their own directory, so drop the same folder there instead. - Ask your agent to use it.
Restart the agent if it was already running, then it picks the skill up with no config.
skills folder
Copy
~/.claude/skills/loyalty-tier/
SKILL.md
references/tier-model.md
designs/
# one file per design, with the net per tier
# macOS: create the folder and open it in Finder
mkdir -p ~/.claude/skills && open ~/.claude/skills
# Windows: paste in the Explorer address bar
%USERPROFILE%\.claude\skills\Frequently asked questions
How is this different from copying a competitor's tiers?
Their thresholds fit their customer distribution, which is not yours. This reads where your customers cluster in annual spend and places thresholds just above those points, which is where a threshold actually changes behavior rather than rewarding it.
Which agents does the skill run in?
Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision.
What does it need connected to work?
Order history with customer identifiers over at least a year, product margin, and shipping cost. History of any previous thresholds or programs sharpens the stretch estimates considerably.
Why avoid discount tiers?
Because the arithmetic almost never works: a discount applies to every customer already above the threshold, and that group is usually far larger than the group who stretch. Free shipping and access benefits cost less per qualifying customer and convert comparably.
How many tiers should a program have?
As many as your distribution has clusters, which is usually two or three. A fourth tier at the top with very few members is often worth it for a different reason: the customers there are worth knowing individually, and the reward is attention rather than money.
What is included with the skill?
The skill itself, the tier model reference, the design format, and the reward cost library. It is free and security scanned. We re-publish the ZIP when the model changes, so download it again if your thresholds start shifting.
A tier threshold is a promise priced in margin.
Polar carries lifetime value and contribution margin per cohort on one definition, so the reward behind each tier stays affordable at the volume it attracts.
<script type="application/ld+json">{"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How is this different from copying a competitor's tiers?","acceptedAnswer":{"@type":"Answer","text":"Their thresholds fit their customer distribution, which is not yours. This reads where your customers cluster in annual spend and places thresholds just above those points, which is where a threshold actually changes behavior rather than rewarding it."}},{"@type":"Question","name":"Which agents does the skill run in?","acceptedAnswer":{"@type":"Answer","text":"Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision."}},{"@type":"Question","name":"What does it need connected to work?","acceptedAnswer":{"@type":"Answer","text":"Order history with customer identifiers over at least a year, product margin, and shipping cost. History of any previous thresholds or programs sharpens the stretch estimates considerably."}},{"@type":"Question","name":"Why avoid discount tiers?","acceptedAnswer":{"@type":"Answer","text":"Because the arithmetic almost never works: a discount applies to every customer already above the threshold, and that group is usually far larger than the group who stretch. Free shipping and access benefits cost less per qualifying customer and convert comparably."}},{"@type":"Question","name":"How many tiers should a program have?","acceptedAnswer":{"@type":"Answer","text":"As many as your distribution has clusters, which is usually two or three. A fourth tier at the top with very few members is often worth it for a different reason: the customers there are worth knowing individually, and the reward is attention rather than money."}},{"@type":"Question","name":"What is included with the skill?","acceptedAnswer":{"@type":"Answer","text":"The skill itself, the tier model reference, the design format, and the reward cost library. It is free and security scanned. We re-publish the ZIP when the model changes, so download it again if your thresholds start shifting."}}]}</script>