Loyalty programs that customers actually use
A Sephora customer can recite their tier from memory. A Starbucks regular knows exactly how many stars away they are from a free drink. That muscle memory is not luck; it's the result of three or four design decisions you can copy. This guide covers the structure, the earn and burn rates, and the five KPIs that tell you the program is working before finance asks.
Key takeaways
- Reward behavior you want more of, not behavior the customer would have done anyway. Discounts on autopilot are not a program.
- Most programs die at the earn-burn ratio, not the creative. Lock that number with finance before you write the launch email.
- Points reward everyone. Tiers reward the top five percent. Most brands need both, in that order.
- Active member share is the metric. Enrolment is a vanity number that looks great in board decks and means almost nothing.
- Liability is a real entry on the balance sheet. Expiry, breakage, accruals: plan them before launch, not after the audit.
A working program rewards behavior the brand wants more of, not the behavior the customer would have done anyway.
Get the earn rate, the burn rate, and the tier story right and the program runs itself for years. Get one of them wrong and you spend three quarters quietly walking it back.
Definition
What a loyalty program actually is
Strip the marketing language. A loyalty program is a contract: the customer agrees to come back more often, and you agree to give them something they value in return. Sephora's contract is points plus tier perks. Starbucks's contract is stars for free drinks. Amex's contract is points for travel. Every working program is some version of that handshake.
Plain definition
A loyalty program is a long-running system that issues points, status, or perks in exchange for repeat behavior, usually purchases or visits, with the goal of increasing customer lifetime value above the cost of the program.
Who runs this
CRM and retention teams own the program. Finance owns the liability. Product owns the surface where customers see their points and tier. The marketing team is the daily operator.
How it differs from adjacent mechanics
- vs promotional discounts. A discount is a one-time price cut. A loyalty program is a relationship that earns its discount through behavior.
- vs subscriptions. Subscriptions charge for access. Loyalty programs reward usage. Many brands run both, with the loyalty layer making the subscription stickier.
- vs gamified campaigns. A campaign is time-bound. Loyalty is always on. Campaigns ride on top of the loyalty layer to spike participation in specific windows.
- vs VIP and concierge programs. VIP is a manual relationship for the top 0.1 percent. Loyalty is a programmatic version that scales the principle to the top 5 percent.
The four signals
A loyalty program only works when all four signals are visible at once.
Four moving parts. Get any one of them wrong and the program reads as decoration.
Earn
How fast does the customer accumulate?
Burn
What can they spend it on?
Tier
Does status feel earned?
Visible state
Where do they see all of it?
Sephora makes earn fast, burn obvious, tier explicit, and state visible on every screen. Most teams ship two of the four and wonder why the program is flat.
The shift
Why discounts run dry and programs compound
The customer is loyal to the price, not the brand.
Each discount works on the customer in front of you. None of them stays after the cart is closed.
Each visit deposits something the customer can come back to.
Points carry forward. Tier carries forward. Status carries forward. The compounding is the program.
Structures
Points, tiers, and perks. Pick the right combination
Three building blocks. Almost every working program uses two of them. Sephora runs points plus tiers. Starbucks runs points only. Amazon Prime runs paid loyalty with perks. The right pick depends on category margin and what your customer is actually buying.
| Structure | Best for | Watch out for | Effort to run |
|---|---|---|---|
| Points only | High-frequency categories (groceries, fuel, food delivery, fashion). Customer needs a new earn event every visit. | Liability balloons quietly. Set expiry and a maximum balance. | Low to medium |
| Tiers only | Status-driven categories (airlines, hotels, premium retail). Customer is buying recognition more than rebate. | Tier inflation. Once everyone is Gold, no one is Gold. | Medium |
| Points plus tiers | The default for most retail and consumer brands. Points reward every purchase, tiers reward the top 5 to 10 percent. | Two parallel economies need consistent communication or the customer gets confused. | Medium to high |
| Tiers plus perks (no points) | Premium and luxury, where points feel transactional. Tiers give status, perks give surprise and delight. | Hard to explain in an ad. Needs a strong onboarding moment. | Medium |
| Paid loyalty | Annual fee for guaranteed perks (free delivery, member pricing, early access). Works when the perks pay for themselves in 1 to 2 visits. | If the perks are not used, churn is brutal at renewal. | High |
Why it matters
The math finance is going to ask about
The working band for the share of revenue a healthy retail loyalty program gives back. Below it, the program is decorative. Above it, margins suffer. Lock the number with finance before designing the catalogue.
Industry working ranges; varies by category and margin
Reward economics
Set the earn and burn rates before anything else
There is exactly one number you cannot get wrong: the share of revenue you give back. One to five percent is the working band for retail. Above that, margins suffer. Below it, the program is decorative. Decide the number with finance in week one, defend it through every creative review, design backwards from it.
Earn rate
How fast the customer accumulates points. Common in retail: 1 point per rupee, 1 point per dollar. The percentage you give back is what matters, not the point label.
Burn rate
What customers spend points on. The exchange rate decides perceived value. 100 points for a 50-rupee discount is a 0.5 percent reward; 100 points for free shipping can feel like 10 percent.
Reward percentage
Earn rate times burn rate. Most retail loyalty programs run 1 to 5 percent. Subscription brands can go higher because retention compounds. Set a target and govern to it.
Liability and breakage
Outstanding points are a liability on the balance sheet. Breakage (points that expire unused) is the offsetting credit. Most programs assume 15 to 30 percent breakage; verify quarterly.
What finance asks for
The three numbers that decide if the program is healthy
A loyalty program is an issuance, redemption, and liability problem dressed as a marketing one. These are the numbers a CFO will ask for in the first meeting after launch.
The economics behind the points
Earn, burn, breakage: the three numbers that decide the program
Points issued (30d)
8.4M
+11%Points redeemed (30d)
5.7M
68%Open liability
$92K
stableIssued, redeemed, and outstanding. Healthy programs run redemption at 50 to 75 percent of issuance.
Worked example
What the math looks like in practice
Pin this to your loyalty brief. Plug in the numbers from your category and the panel tells finance the annual cost before launch.
Annual cost of a 1-point-per-rupee program
Worked example for a retail brand running 250 crore in annual revenue, 1 point per rupee earn rate, 100 points = 50 INR redemption value, 65 percent redemption rate.
Inputs
Annual revenue
2,500,000,000 INR
Trailing 12 months
Earn rate
1 point per 1 INR
Base rate, no multipliers
Redemption value
100 pts = 50 INR
0.5 INR per point
Redemption rate
65%
Healthy for retail
Effective reward percentage
revenue × 0.5/100 × 0.65 = annual reward cost
Output
Annual reward cost
8,125,000 INR
0.65% of revenue
Outstanding liability
~12,500,000 INR
Unredeemed points carried
Breakage credit
~3,500,000 INR/yr
20% expiry assumption
Net P&L impact
~4,625,000 INR
Reward cost minus breakage
Best practices
The seven rules of a program that lasts
- 1
Make the first reward easy to reach
If the first redeem-able reward is more than 3 to 5 visits away, most members never see it. Front-load a small reward at signup or after the first purchase. The first redemption is the strongest predictor of long-term loyalty.
- 2
Show the next milestone, always
Every member-facing surface should answer one question: how close am I to my next reward or tier. A progress bar with a target does more for engagement than a richer reward.
- 3
Communicate balance and expiry on every channel
Email, SMS, push, account page, receipt. Members forget they have points. Reminders before expiry produce a measurable redemption spike and protect breakage assumptions.
- 4
Tier with care, not by accident
Set tier thresholds at the 80th, 95th, and 99th percentile of customer value. Re-qualify each year. Avoid hand-issuing tiers; once the rule breaks, every customer-service request becomes a negotiation.
- 5
Let perks surprise, not points
Predictable points keep the program transactional. A small unexpected perk on a top-tier birthday or anniversary is the part members talk about. Reserve 5 to 15 percent of program budget for surprise.
- 6
Run quarterly health checks
Watch active member share, redemption rate, and tier inflation each quarter. Adjust earn rates by region or product category if the economics drift, but communicate every change clearly.
- 7
Make leaving graceful
Members will go inactive. A respectful winback (we miss you, here is a small bonus) outperforms aggressive expiry threats and keeps the brand from feeling punitive.
Use cases
When a loyalty program is the right call
High-frequency retail
Grocery, pharmacy, fast fashion, food delivery, fuel. Customer purchases 6 or more times a year.
Points-plus-tier programs typically lift annual frequency 8 to 15 percent and shift category share.
Subscription and digital
Streaming, SaaS, fintech, mobility. Loyalty layers on top of the subscription to reward depth, not just renewal.
Renewal lifts 3 to 8 percentage points. Cross-sell into adjacent products improves once tier perks unlock them.
Hospitality and travel
Airlines, hotels, OTAs. Tier-driven status with experiential perks (lounge access, room upgrades).
Top-tier members deliver disproportionate revenue and brand advocacy. Most-loved category for loyalty design.
Premium and beauty
Cosmetics, fragrance, accessories. Tier-plus-perks with sample-size rewards and curated drops.
Repeat rate lifts and review velocity climbs because top-tier members are the brand evangelists.
When to skip
Loyalty programs are not a fit when...
Purchase frequency is once a year or less
Furniture, appliances, weddings. The reward never gets earned in time to matter. A referral program or a post-purchase service program is a better fit.
Margin cannot support a 1 to 5 percent rebate
Some categories (commodities, low-margin marketplaces) cannot give back a meaningful share. A status-only or perks-only program may still work, but points will fail quietly.
The brand is in a sensitive purchase context
Insurance claims, healthcare, debt repayment. Points-and-tiers framing reads as flippant. Service excellence and personalization beat gamified loyalty.
Operations cannot honor the rules consistently
If franchisees, third-party stores, or partner platforms cannot redeem points reliably, the program creates more support tickets than goodwill. Fix the operations layer first.
Common mistakes
The mistakes that quietly kill programs
Mistake
Over-rewarding the first purchase. Welcome bonuses so generous that profit per new member goes negative.
Fix
Cap welcome rewards at the value of the second expected purchase, not the first. Move the rest of the budget to tier upgrade rewards where retention is already proven.
Mistake
Tier inflation. Eighty percent of members reach Gold within a year and the tier loses meaning.
Fix
Set tier thresholds at percentile (top 10 percent for Gold, top 1 percent for Platinum) and re-qualify annually. Add a softer 'maintain' rule so churn does not feel punitive.
Mistake
Hidden balance. Customers do not know they have points until they get an expiry email.
Fix
Show balance in the header of the account page, on every order receipt, and in monthly summaries. Surface the next reachable reward in the same view.
Mistake
Confusing earn rules. Different rates for different categories that members cannot remember.
Fix
Default to one simple base rate. Use bonus events (2x weekends, 3x category Tuesdays) for variation. Consistency beats cleverness.
Mistake
No exit ramp. Inactive members stay enrolled forever, polluting the data.
Fix
Define inactivity (no earn or burn for 12 to 18 months) and run a winback. Members who do not reactivate move to a 'dormant' state with limited communication.
Measurement
The five KPIs that prove the program works
| KPI | What it measures | Healthy range |
|---|---|---|
| Active member share | Members who earned or burned in the last 90 days, divided by total enrolled. | 40 to 65% |
| Member revenue share | Revenue from active members divided by total revenue. The clearest signal that loyalty is concentrating value. | 55 to 80% |
| Redemption rate | Points redeemed divided by points earned in the period. Low redemption signals breakage but also disengagement. | 55 to 75% |
| Repeat purchase frequency lift | Member purchase frequency vs matched non-member cohort. The cleanest test of incremental impact. | +10 to +25% |
| Reward liability ratio | Outstanding point value divided by trailing 90-day revenue. Stable ratio means the program is in balance. | 1 to 4% |
Pre-launch checklist
What to confirm before you ship the program
Three programs you already know
What each one is buying with the design choices it made
Same word, very different programs. Read these like a marketer reads a competitor analysis: the choice is the playbook.
Sephora Beauty Insider
Three tiers (Insider, VIB, Rouge) with point rebates, exclusive drops, and birthday gifts.
Top tier is 1 to 3 percent of customers and a much larger share of revenue. The aspirational tier sells the program to the middle.
Starbucks Stars
Single-currency points (Stars) redeemable for free drinks and food. No tiers. Bonus star challenges twice a month.
Mobile-first. The app is the loyalty surface, which makes the program a daily marketing channel and not just a perks page.
Amex Membership Rewards
Points earned on spend, redeemable for travel and partner perks. Tiered cards (Green, Gold, Platinum).
The card itself carries the status. Points fund the catalogue; the tier carries the relationship. Premium loyalty done as a paid product.
Sweetgreen Sweetpass
Paid loyalty: $10 a month for $3 off every order. Plus a free-tier program with points.
Paid loyalty when the perks pay for themselves in 1 to 2 visits. Locks in frequency the way a subscription does, while keeping the door open via free tier.
Design rules
What separates a program that lasts five years from one that quietly sunsets in eighteen months
Good loyalty programs feel generous, not transactional.
Make the first reward easy to reach
If the first redeem-able reward is more than 3 to 5 visits away, most members never see it. Front-load a small reward at signup or after the first purchase.
Show the next milestone, always
Every member-facing surface should answer one question: how close am I to my next reward or tier. Progress bars do more work than richer rewards.
Communicate balance and expiry on every channel
Email, SMS, push, account page, receipt. Members forget they have points; reminders before expiry produce a measurable redemption spike.
Re-qualify with grace
Members lose status because life happens, not because they stopped caring about the brand. A short grace period and a clear path back saves the relationship.
The most loved loyalty programs do not feel like "spending money to get something back." They feel like a brand quietly remembering what the customer keeps coming back for.
That is why the mechanics matter less than the feeling. Sephora does not win because of points. Sephora wins because the program reads as a relationship that compounds, not a discount queue that resets.
In the wild
Three programs and what they did right
Beauty retail
Three tier program (Insider, VIB, Rouge) with point rebates plus exclusive product drops and birthday gifts.
Outcome. Top tier represents 1 to 3 percent of customers and 20-plus percent of revenue. The aspirational tier sells the program.
Coffee chain
Stars program: every purchase earns stars, redeemable for free drinks. Bonus star challenges twice a month.
Outcome. Mobile order share lifts because the app is the loyalty surface. Daily-active behavior turns the program into the marketing channel.
Airline
Frequent flyer miles plus elite status (Silver, Gold, Platinum) with re-qualification each year.
Outcome. Top-tier members carry the route economics. Status retention drives behaviour beyond price for the highest-yield segment.
Implementation
Build this with Bricqs
Bricqs gives you the points engine, tier rules, reward inventory, and member-facing surfaces in one place. Configure it from the dashboard or wire it into your existing app via the SDK and APIs.
Frequently asked
Common questions before launch
How long does it take to launch a loyalty program?
A working v1 with points, one tier, and a reward catalogue can ship in 2 to 4 weeks on a builder. Full integration with POS, CRM, and finance reporting usually takes 2 to 3 months. Most teams launch the soft version first and harden it after.
Should we charge for loyalty membership?
Only if the perks pay for themselves in 1 to 2 visits. Free shipping, member pricing, and early access are the most common paid-loyalty perks. Free programs are still the default for mass-market brands.
How do we handle points after a refund?
Reverse points proportional to the refund and log it in the same ledger. Bricqs and most loyalty platforms support automatic reversal on order cancellation. Document the rule in the public terms so customer-service teams can quote it.
What is a typical reward percentage?
Retail: 1 to 3 percent on the base rate, with bonus events pushing weighted average to 2 to 5 percent. Subscription: 5 to 10 percent because retention compounds. Premium: lower base, higher tier perks. Always model annualized cost before launch.
Should we let points expire?
Yes, with notice. 12 to 24 month expiry from last earn event is standard. Send reminders 60, 30, and 7 days before expiry. Expiry without notice damages trust more than it saves liability.
How do we re-launch a tired program?
Audit the earn-burn ratio, simplify the rules, add a visible progress UI, and ship a Welcome Back campaign for dormant members with a small bonus. Most relaunches are about clarity and visibility, not new mechanics.
Branch by goal
Where to go from loyalty
Loyalty is the always-on layer. Most teams pair it with a campaign mechanic on top. Pick the next read based on what you are trying to move.
If your goal is
Drive repeat frequency
Visible progress, points multipliers, milestone perks. Layer them on the loyalty spine.
Read the playbookIf your goal is
Build a daily habit
Streaks layer cleanly on top of points and tiers. The freeze keeps them humane.
Read the playbookIf your goal is
Run a seasonal push
A festival or launch sits on top of loyalty without breaking the always-on contract.
Read the playbookIf your goal is
Recruit through members
Top-tier members are your cheapest acquisition channel. Build the referral program around them.
Read the playbookIf your goal is
Tune the points economy
Earn rates, spend menus, multipliers, caps. Get the math right before scaling.
Read the playbookIf your goal is
Refresh the reward catalogue
Stale catalogues are why programs go quiet. Quarterly rotation, three to five reward types.
Read the playbookReady to ship?
Try Bricqs free, or talk to a strategist
Plan a campaign, configure the engine, and ship in days. No credit card required to start.