Shopify Loyalty Program: Step-by-Step Setup for Plus Teams

The programme design here is what separates a loyalty scheme that pays for itself from one that hands margin back on orders that would have happened anyway. Most guides to a Shopify loyalty program stop at “install an app and pick a points rate”. This one covers the settings you actually have to decide, the order to decide them in, and the step most teams get wrong: defining which revenue earns points before anyone has priced the liability.

The guide is written for operators at $3M–$30M in revenue on Shopify Plus or a paid subscription platform, where repeat orders carry a meaningful share of revenue. If you are below that floor, a points scheme is rarely your first lever and the setup here will cost more attention than it returns. If you are choosing software rather than designing the programme, start with the loyalty platform roundup instead.

What does a Shopify loyalty program actually need to do?

A Shopify loyalty program needs to change a specific customer behaviour at a cost you can state in advance. That is the whole test. It is not a rewards catalogue, a badge system or a place to park a birthday discount.

Shopify’s admin does not ship a complete points-and-tiers engine, so a loyalty program on Shopify is nearly always a combination of a loyalty app, the customer account pages, store credit or discount codes, and your email platform. Each of those is a separate system with its own rules. The design work is making them agree.

The failure mode looks like a leak. Points are issued on every order, a slice of them redeem, and the redemptions land on customers who were already about to buy. Margin drains out through the discount code and nothing measurable comes back. The steps are ordered so that leak is priced before it is built.

Prerequisites before you touch an app

Three things need to exist first, and the programme is guesswork without them.

1. A gross margin figure per order, or per product group. Not revenue. Every rate you set derives from it.
2. A repeat-purchase baseline. The share of customers who order a second time, and how long it takes. If you cannot read this from Shopify’s customer reports or your warehouse, fix that first. The subscription churn calculator helps if part of your repeat revenue is recurring.
3. An owner for liability. Someone in finance who agrees that outstanding points are a real obligation and looks at the balance monthly.

If you sell on subscription, also have the cancellation reasons to hand. Why supplement subscribers cancel at month three shows the kind of timing pattern worth knowing before you decide when a reward should land.

How do you set up a Shopify loyalty program step by step?

The seven steps run in the order the decisions depend on each other. Skipping ahead to the app configuration is the usual mistake, because the earn rate in the app is only a number you typed in unless the earlier steps produced it.

Step 1: Give the programme one job and one margin budget

Write one sentence naming the behaviour. “Move first-time buyers to a third order inside the repurchase window” is a job. “Increase loyalty” is not.

Then set the margin budget: the share of gross profit per order the programme is allowed to spend. There is no published correct figure, and any number quoted without your margin is a metric to confirm. What you do have is a method. Take gross profit per order, decide what fraction you would happily give back to secure a repeat order, and treat that as a hard ceiling on the combined cost of earn plus redemption.

The job statement also decides later choices, such as whether tiers are built on spend or on order count. A consumable brand paying for frequency should not build its top tier on basket size.

Step 2: Choose the mechanism your platform can enforce

There are four common mechanisms, and they are not interchangeable.

Take from the table that points and tiers give the most control, and store credit gives the most trust. Perks alone are the cheapest to try but the hardest to measure.

Whichever you choose, check that the app can enforce the earn, refund and renewal rules this guide sets out. Ask the vendor for the setting names, not a feature list. Whether a given app supports earn-on-amount-charged, refund reversal, and renewal handling is documented in its own help centre, and it varies by app and by plan. Confirm it there before you commit. The Yotpo pricing and plans piece shows the packaging shape to expect when a loyalty module is bundled with reviews.

Step 3: Derive the earn rate and the redemption value

Do the arithmetic backwards from the margin budget. Two settings matter: the earn rate (points per currency unit spent) and the redemption ratio (how many points buy how much value).

Here is a hypothetical, illustrative case with made-up round numbers: an order of $60 with gross margin of 55% carries $33 of gross profit. Say the programme pays 1 point per $1 and 100 points redeem for $5. That illustrative $60 order earns 60 points, worth $3 at full redemption. That is about 9% of the order’s gross profit if every point is redeemed. Tighten or loosen either setting and the percentage moves in a way you can see before launch.

Full redemption is the ceiling, not the forecast. Some balance is never redeemed (the industry word is breakage), and you should plan on your own redemption data rather than an assumed rate. Until you have run for a full cycle, model the ceiling and treat any lower estimate as a metric to confirm.

Three related settings deserve an explicit decision instead of the default:

• Minimum redemption threshold. A minimum balance to redeem stops many tiny redemptions and nudges customers toward another order to reach it.
• Redemption cap per order. A cap stops one large balance from wiping out the margin on a single basket.
• Expiry rule. Inactivity-based expiry (balance lapses after no activity) usually feels fairer than a fixed calendar date. Some regions regulate expiring balances, so confirm specifics with counsel.

Step 4: Set the earn base and the refund reversal

The earn base is the step most teams get wrong, and the one that turns a sensible programme into a margin leak.

The earn base is the definition of which revenue earns points. Defaults in many apps pay on the order subtotal before discounts, sometimes including shipping and tax. In a hypothetical case, a customer uses a welcome code worth a fifth off a $60 order and the programme pays points on the pre-discount $60, so you have discounted the order and then rewarded the full amount. Pay on the amount the customer was actually charged for products, excluding tax, shipping and gift cards, unless you have a specific reason not to.

The reversal rule is the second half. Points earned on an order must be removed when that order is refunded, partly refunded or cancelled. Without reversal, a refund returns the money and keeps the points, and a customer who orders and returns repeatedly is paid for it. Check in your app’s settings that reversal is on, and test it with real orders rather than trusting the label.

Then look at the edge cases, because they are where balances go wrong:

• Exchanges: does the swapped item earn again, or does the original earn carry?
• Partial refunds: do points scale down proportionally?
• Orders paid partly with points: do the redeemed points count toward the earn base? Usually not.
• Orders edited after purchase: does a post-purchase upsell earn?

Write the answer to each in one line. That short document is what your support team will reach for when a customer asks why the balance changed.

Step 5: Write the rules for subscribers and discount stacking

Subscribers are the awkward case, and for a brand whose repeat revenue is recurring, the first place the programme collides with the rest of the business.

A subscription renewal is an order the customer did not decide to place this month. Paying full points on it rewards inertia. It also usually already carries a subscriber discount, so the same stacking problem as the earn base appears every cycle.

There are three defensible rules, and the choice depends on the job you wrote for the programme:

1. Renewals earn on the amount charged. Simple, and subscribers keep feeling rewarded.
2. Renewals earn nothing, subscribers get a different perk. Cleaner accounting, and it keeps the points economy for one-time buyers.
3. Renewals earn a reduced rate. More to explain, more settings to maintain.

Whichever you pick, put it in writing and check that your subscription app and your loyalty app both apply it. They are separate systems. A renewal created by Recharge or a similar platform reaches Shopify as an order, but whether the loyalty app treats it as an earning order depends on how it reads the order source. That behaviour is documented per app, so check it in the settings rather than assuming.

Point-of-view worth stating: a loyalty scheme is a poor tool for reducing involuntary churn. If a subscriber leaves because a card failed, a points balance does nothing. Failed payments are a payment-recovery problem, covered in how to reduce involuntary churn. Loyalty helps where the customer is choosing to stay or leave. If you sell on subscription, the best subscription app comparison for Shopify covers how the billing side hands data to the store.

Step 6: Connect points to Klaviyo flows and the storefront

A programme customers cannot see does not change behaviour. Two connections do most of the work.

The customer account. Show the balance, the next reward and what is needed to reach it. This is a display setting in most apps and a theme edit in others. Put the balance somewhere a customer sees it on the way to checkout, not only in the account area they rarely open.

Your email platform. The loyalty app needs to send balance, tier and last-earn date to Klaviyo as profile properties or events, so flows can use them. Typical uses:

• A points-expiring reminder sent before a balance lapses, which turns liability into an order.
• A “you’re one order from the next reward” message triggered by the gap to a threshold.
• A tier-change message when a member moves up or drops.
• A suppression rule so points messages do not fire on top of a win-back or abandoned cart flow at the same moment.

The last item is where teams create noise. Give each of the flows a priority and a frequency cap, or a customer who has a balance, a cart and a subscription renewal can receive three messages in a day. See Klaviyo flows for how to order and cap triggered messages.

Then check the sync direction. If Klaviyo reads points from the app, a delay in the sync means a customer is told they have a balance they no longer have. Find out from the app’s documentation how often it pushes updates, and design the email copy so a short lag is harmless.

Step 7: Test with real orders before launch

Do not launch on a settings review. Use a small set of staff or test accounts and place real orders on the live store, then run the awkward cases through them.

Work through this list and note the balance after each action:

1. A plain order with no discount. Does it earn the amount you calculated in Step 3?
2. An order with a discount code. Is the earn on the amount charged?
3. An order that includes shipping and tax. Are both excluded from the earn base?
4. A full refund. Do the points come back off?
5. A partial refund. Do they scale?
6. A redemption, followed by a refund of the order that used it. Does the redeemed value return to the balance, or vanish?
7. A subscription renewal. Does it follow the renewal rule you wrote?
8. A cancellation before fulfilment.

Any mismatch is a rule you thought you had set but had not. Fix it, repeat the test, and only then announce the programme. A wrongly configured earn base is far cheaper to fix before the first customer has a balance than after, when changing it means either breaking a promise or absorbing the cost.

What goes wrong after launch, and how do you catch it?

Most of the trouble surfaces in the first full repurchase cycle. Three things break repeatedly, and each has a signal you can read.

Liability drifts up faster than orders. If outstanding points grow while repeat orders do not, the programme is issuing rewards without changing behaviour. Read the balance against gross profit monthly. The signal is the ratio moving, not either number alone.

The programme rewards people who would have bought anyway. Members are self-selected loyal customers, so their revenue looks good whether or not the scheme did anything. The only clean read is a holdout: a group of comparable customers who do not receive the programme, compared on repeat-purchase rate and time between orders. Without one, any “members spend more” claim is a metric to confirm.

The rules are documented nowhere. A customer writes in asking why 30 points vanished after a refund, and support guesses. Publish a plain-language rules page and give support the one-line answers to the earn-base edge cases.

If you want a benchmark for how a consumables brand’s subscription cohorts decay before deciding what reward timing to test, the DTC consumables churn benchmark is the reference to use, with its sources.

Which loyalty designs are not worth building?

A few designs are common and rarely pay back, and each has an honest exception.

• A points scheme for a product bought once every few years. There is no second order to influence. A referral or review mechanism fits better. This is the clearest case for being told plainly that the programme is not for you.
• A complex tier ladder on a small customer base. Each tier needs enough members to matter, a distinct perk and a cost you have modelled. Two tiers you can explain beat five you cannot.
• Points as a substitute for a good subscription offer. If subscribers are cancelling because of product fit or delivery timing, a balance will not hold them.
• Rewards for actions that do not spend money. Points for a social follow or a review cost little each, but uncapped they invite gaming and put a liability against customers who never buy.

The exception to the last one is reviews. A review incentive is worth having when it is capped, tied to a verified purchase and paid modestly, because the review has its own value on the product page. Compare tools such as Trustpilot and Yotpo before choosing where reviews and loyalty should live.

How much ongoing work does a loyalty programme take?

Less than a full flow rebuild, more than “set and forget”. The recurring jobs are small and specific:

• A monthly liability read against gross profit, owned by finance.
• A quarterly check of the redemption ratio against the margin budget, because product costs and shipping costs move.
• A review of flows that use loyalty data, so a tier change or an expiry reminder still fires correctly after theme or app updates.
• A review of the rules page after any change to the subscription or refund policy.

The hidden line items are worth pricing: the app’s own fees (packaging varies by vendor and by plan, so read the current pricing page), staff time for the monthly reads, and the discount cost itself, which is the largest of the three and the only one that scales with success.

The programme stops being worth it when the measured lift over a holdout is smaller than the cost of the rewards paid to members who would have ordered anyway. If you cannot show a holdout result, you do not yet know which side of that line you are on.

How does loyalty fit with the rest of retention?

A loyalty program is one lever in a retention system, and it is the wrong lever for two of the three main leaks. Voluntary churn from a customer who no longer wants the product needs an offer, a pause option or a product fix. Involuntary churn from failed payments needs recovery on the billing side. Loyalty helps in the third case, the customer who is drifting, has options and needs a reason to place the next order.

Treating loyalty as one lever changes what you build. You size the reward to the drifting customer, time it to the point in the cycle where drift starts, and measure it against a holdout rather than against total member revenue.

Teams that reach this point usually find the loyalty rules, the subscription rules and the email logic have been configured by three different people in three different tools. Reconciling them is a subscription retention problem: one set of definitions for what a renewal is, what it earns, and what a cancellation reverses. Pointerflow’s subscription retention service covers that reconciliation across Shopify, the subscription platform and Klaviyo.