Which loyalty program KPIs should you track?
Loyalty program KPIs measure whether eligible customers join, use benefits, return to buy and renew, and whether the program’s costs are justified. Start with enrollment, first-benefit activation, reward redemption, repeat purchasing, paid renewal, reward cost and purchase frequency. A larger member count alone does not show that the program caused profitable growth.
Choose a small set that supports actual decisions. For a new launch, focus on joining and first use. For a mature program, add repeat purchase, renewal and contribution analysis. A free program does not need a paid-renewal KPI.
This guide covers loyalty measurement after launch. Use the ecommerce metrics guide for store-wide KPIs and the business-case worksheet for investment planning. The distinction between program health and financial performance is also useful in Oracle’s loyalty KPI framework.
Download the monthly loyalty KPI worksheet
The free Excel workbook includes definitions, twelve review periods, seven calculated metrics per period and a fictional example. Enter your numerator, denominator, source, cohort and next action. The CSV version provides the same blank monthly rows for manual calculation. No email address is required.
This is a manual planning and reporting worksheet, not a live Memberply dashboard or a Shopify connector. It does not calculate causal lift or automatically judge whether performance is good.
Blue cells are editable; lavender cells calculate results. Blank inputs remain blank. Enter a verified zero when nothing happened. A zero denominator or invalid count displays “Check inputs.” Leave inapplicable metrics blank and record why. Choose one currency and document exclusions before entering values.
Seven loyalty metrics with formulas and examples
On small screens, scroll the table horizontally.
| Metric | Formula | Example | Scope rule |
|---|---|---|---|
| Enrollment rate | Eligible exposed nonmembers who joined ÷ Unique eligible nonmembers exposed to the offer | 40 ÷ 200 = 20% | Use the same invitation cohort and attribution window; exclude existing members. |
| First-benefit activation rate | New members who used a defined benefit within 30 days ÷ New members with a full 30 days of observation | 24 ÷ 40 = 60% | Choose one qualifying benefit action; joining and logging in alone do not count. |
| Reward redemption rate by value | Value redeemed from the selected issuance cohort by cutoff ÷ Original issued value of that same reward cohort | 300 ÷ 600 = 50% | Use one currency or points unit; exclude refunds and adjustments from redemption; record reversals separately. |
| Member repeat-purchase rate | Members in the cohort with a second qualifying order within 90 days ÷ Members in the first-purchase cohort with a full 90 days of observation | 18 ÷ 60 = 30% | Exclude canceled/test orders; hold the cohort fixed and use the same window for comparisons. |
| Paid membership renewal rate | Memberships due in the period successfully renewed by cutoff ÷ Memberships due to renew in that period | 45 ÷ 50 = 90% | Count memberships consistently; record grace period and late-payment cutoff. Not applicable to free tiers. |
| Reward cost rate | Reward costs allocated to the review period ÷ Eligible net product sales in the same scope and period | 420 ÷ 12,000 = 3.5% | Use a consistent costing policy and currency; exclude tax, shipping and gift-card sales from this product-sales basis. Avoid counting the same cost twice. |
| Orders per purchasing member | Qualifying member orders in the period ÷ Distinct members with a qualifying order in that period | 90 ÷ 60 = 1.5 orders | This excludes members who did not buy; it is purchase frequency among buyers, not all members. |
Multiply a ratio by 100 when expressing it as a percentage. The workbook applies percentage formatting to the first six metrics. Orders per purchasing member remains a number. These definitions are working choices, not claims that every reporting product uses the same formula.
Enrollment measures conversion among eligible people who actually saw the offer. If exposure cannot be measured reliably, report new joins as a count rather than dividing by unrelated store traffic. Activation and repeat purchasing need mature cohorts: someone who joined yesterday has not had a full 30-day opportunity to activate.
Renewal is measured among memberships due, not every active member. Record whether success means a completed payment by the due date or by an agreed grace-period cutoff. Keep scheduled cancellation, failed payment and inactive status distinct.
Measure redemption using the same reward cohort
Suppose a store issues 600 units of reward value to a defined group in January. By the end of March, customers have redeemed 300 units from those exact issuances. The cohort’s value redemption rate is 300 ÷ 600 = 50%. Record the issuance dates, cutoff, currency or point unit, and any reversals or adjustments.
If customers redeem 300 units in March while the store issues 200 new units in March, 300 ÷ 200 = 150% is a flow ratio, not the redemption rate of the March issuance cohort. Earlier rewards can be spent later. Do not label that flow ratio as the percentage of new rewards used.
Count unique redeemers separately if you want participation by person. One member redeeming several times should not become several people. A value-based rate and a member-based rate answer different questions.
Low redemption can mean poor awareness, hard-to-use rewards or simply an observation window that is too short. It does not automatically mean the rewards are unwanted. Nor is unredeemed value automatically profit: outstanding obligations, expiry and cost recognition need a consistent finance policy.
Build a reliable measurement record
On small screens, scroll the table horizontally.
| Source | What to collect | What to verify |
|---|---|---|
| Membership records | Customer identity, tier, joining date, effective access dates, renewal due dates and payment outcomes. | Use status at the event time where needed. Today’s active-member list does not reconstruct historical membership. |
| Orders | Customer ID, order date, qualifying sales, cancellations, returns and channel. | Deduplicate customers, exclude test activity and apply one sales definition consistently. |
| Reward records | Issuance identifier, value, currency, redemption, reversal and expiry events. | Join redemption to the original issuance where the cohort metric requires it. If unavailable, label the metric unavailable. |
| Offer and benefit events | Eligible offer exposure, joins and the chosen benefit-use action. | Define what counts as exposure and activation. Page views and login events are not automatically benefit use. |
| Cost records | Allocated rewards, app fees, support, promotion and additional fulfillment costs. | Keep costs in the same period and avoid counting expense already deducted from sales or contribution twice. |
Shopify’s customer reports include customer and cohort views, but their time scopes matter. The default cohort grouping is based on first order, which differs from joining a membership. Reconcile a small sample before using a report as your membership denominator.
Memberply’s documented Klaviyo integration syncs membership profile properties such as status and tier. That does not establish that reward balances, redemptions or your full order history are included in that sync. Identify a verified source for each KPI instead of assuming one integration supplies everything.
Start with the implementation checklist to test the customer journey, then record how each observed action reaches the measurement source. This guide does not promise that every input can be exported from one built-in screen.
Separate member spending from incremental contribution
Members may already be a store’s most frequent buyers. Comparing all members with all nonmembers can exaggerate the program’s impact because the groups differ before joining. A rise in member revenue also can reflect seasonality, promotions or a larger member base.
Where feasible, use a well-designed eligible holdout or compare groups with similar prior behavior and equivalent observation windows. Record remaining differences. A matched comparison can inform a decision, but it does not remove every source of bias. Use the cohort analysis guide to keep group and time definitions explicit.
Here is a fictional contribution example. An offer group of 100 eligible customers produces $8,000 in net product sales, compared with $7,000 for an equally sized comparable group. The observed difference is $1,000. At a 40% contribution margin before program costs, that corresponds to $400. Subtract $250 of program costs not already included in the margin, and the estimated net contribution difference is $150.
The arithmetic does not prove the offer caused the difference. Keep selection bias, small samples, refunds and delayed purchases visible. Allocate program costs to the measured scope and avoid subtracting discounts or reward expense twice. Fee-funded credit needs separate treatment of the value owed to members; membership cash receipts are not automatically profit.
Use each result to choose a useful next action
On small screens, scroll the table horizontally.
| Pattern | Investigate | Possible next test |
|---|---|---|
| Few joins despite offer exposure | Audience fit, price clarity, benefit relevance and joining friction. | Test a clearer offer explanation with a comparable eligible audience. |
| Joins rise but first use stays low | Account access, benefit discoverability and eligibility confusion. | Observe a new member locating and using one benefit. |
| Rewards are issued but little is redeemed | Reward availability, cohort maturity, checkout steps and restrictions. | Test the exact redemption path and explain one verified available benefit. |
| Purchase frequency looks high | Whether the denominator excludes non-buyers and whether a few heavy buyers dominate. | Report cohort repeat purchase alongside buyer-only frequency. |
| Renewal falls | Failed charges, voluntary cancellation and missing perceived value. | Separate payment recovery from benefit-value improvements. |
| Sales grow while contribution weakens | Reward cost, discount stacking, support and shipping subsidies. | Review the benefit mix against the business case before increasing incentives. |
Set targets from your own baseline, offer economics and observation window. The worksheet intentionally has no universal red or green threshold. High redemption can be healthy use or an expensive offer; low reward cost can be efficient or evidence that members cannot use the program. Interpret related measures together.
Run a monthly review without shortening the purchase cycle
Freeze the scope
Record the period, cutoff, program, tier, channels, currency and cohort definitions. Keep sources and exclusions beside each result.
Reconcile before calculating
Check sample members and orders against source records. Investigate duplicates, delayed events, returns and missing identifiers.
Compare like with like
Review mature cohorts at the same age. A monthly meeting can evaluate a 90-day purchase window; it does not require every customer to buy every month.
Choose one action and owner
Record the observation, hypothesis, next test, decision date and owner. Keep the prior definition when comparing results over time.
Do not average percentage rows to create an annual result. For genuinely non-overlapping groups with the same definition, recompute from the summed numerators and denominators. Where the same members or reward issuances appear in several snapshots, deduplicate or report the snapshots separately.
For a new program, establish measurement during the pilot. For an existing program, use results to improve onboarding, benefit engagement and renewal communication rather than simply adding more rewards.
Questions merchants ask
What are the most useful loyalty program KPIs?
Start with enrollment, first-benefit activation, reward redemption, repeat purchase, paid renewal, reward cost and orders per purchasing member. Choose the metrics that match your program and pair activity with contribution analysis.
What is a good loyalty redemption rate?
There is no universal target in this guide. Compare the same reward cohort at the same age, then assess usability, costs and the customer purchase cycle. Do not compare different observation windows as if they were equivalent.
How do I calculate loyalty reward redemption rate?
For a value-based cohort rate, divide value redeemed from a selected issuance cohort by the original value issued to that same cohort, measured through a stated cutoff. Use one currency or point unit and record reversals separately.
Does higher member spending prove a loyalty program works?
No. Customers who choose membership may already spend more. Use comparable groups or a well-designed holdout, account for costs and disclose remaining uncertainty before attributing a difference to the program.
Does the worksheet connect to Shopify or Memberply?
No. It is a manual Excel or CSV worksheet. Enter verified inputs from your chosen sources and record definitions, dates and follow-up actions. It does not provide a live dashboard.
How often should I review loyalty metrics?
A monthly review is useful, but each metric needs an appropriate observation window. Compare mature cohorts at the same age and allow enough time for the normal purchase or renewal cycle.