Loyalty Program KPIs: Holdouts Test ROI in Retail and Hospitality

2 days ago

10 min read

A core set of KPIs separates loyalty programs that prove their worth from those running on faith: active rate, redemption rate, retention, repeat purchase rate, CLV, sales penetration, incremental revenue, loyalty ROI, and NPS. The real discipline is organizing them into three tiers, member behavior, program health, and business impact, and testing incrementality with a holdout group before trusting any ROI figure.

TL;DR:

Track behavior, program health, and business impact separately: enrollment and activity move quickly, while retention and financial results take longer to reveal change.
Calculate ROI from incremental margin, not member revenue alone; compare customers receiving offers with a no offer holdout for at least one purchase cycle.
Set prelaunch baselines for retention, average order value, and purchase frequency, then tie improvement targets to margin and test them against reward costs.
Use daily, weekly, monthly, and quarterly reviews for operational flags, campaigns, program health, and executive ROI, while segmenting results and defining each KPI consistently.
Link verified profiles and consistent member IDs to transactions; guest checkout records cannot support individual retention, repeat purchase, or redemption tracking.

KPI quick reference: the core loyalty KPIs and what each tells you
How to calculate the key KPIs: formulas, required data fields, and examples
Program health vs. business-impact framework: what to measure at each tier and how they connect
Measuring incrementality and attribution: holdouts, A/B, and common attribution traps
Benchmarks and target-setting: how to pick KPI targets and interpret performance
Operationalizing measurement: dashboards, segmentation, cadence, and reporting to the C-suite
Analysis of guest profile data and social login metrics as loyalty indicators
Author perspective on measurement priorities for loyalty teams
How MyDigiMenu supports loyalty KPI measurement and program operations
FAQ
Sources

KPI quick reference: the core loyalty KPIs and what each tells you

Every useful loyalty metric earns its place by answering a specific question: are people engaging, is the program working, and is it making money. We find it helps to sort KPIs by which of those three questions they answer, since mixing them is how dashboards turn into noise.

Member behavior KPIs come first because they move fastest and warn you early:

Enrollment rate measures how many eligible customers join; a stall here usually means the sign-up offer or moment is weak.
Active rate tracks the share of members who transact within a set window; a gap between enrollment and active rate points to a dormant-member problem.
One-and-done rate flags members who join and never return, a leading signal that onboarding or early rewards need work.

Program health KPIs tell you whether the mechanics are actually changing behavior:

Retention rate shows whether members keep coming back period over period.
Repeat purchase rate measures how often members buy again versus one-time customers.
Redemption participation tracks the share of members who actually cash in points or rewards, since unused points accumulate as a cost without changing behavior.

Business impact KPIs are lagging by nature but are the ones that justify the budget:

Sales penetration shows what portion of total revenue comes from members.
Incremental revenue and incremental margin isolate the sales the program caused, not just correlated with.
Loyalty ROI ties the whole thing to the cost of running it.

Oracle’s CrowdTwist team frames this same split as health KPIs versus performance KPIs, grouping enrollment, active rate, redemption, tier distribution, and sentiment on one side, and spend per member, sales penetration, and ROI on the other, providing a useful guest experience guide for hospitality pros that complements the NPS and sentiment discussion. The pattern holds regardless of what you call the buckets: engagement metrics lead, financial metrics lag, and you need both.

How to calculate the key KPIs: formulas, required data fields, and examples

Formulas only work when you have the right data fields sitting next to them. Pull together customer ID, order date, revenue per order, member flag, points issued and redeemed, reward cost, and campaign tags before you start, because most calculation errors come from missing one of these, not from the math itself.

Retention rate = (customers active at period end who were also active at period start) ÷ (customers active at period start) × 100.
Churn rate = 100 minus retention rate for the same period.
Repeat purchase rate = (customers with 2+ orders in a period) ÷ (total customers in that period) × 100.
Purchase frequency = total orders ÷ total unique customers over a defined window.
Average order value (AOV) = total revenue ÷ total number of orders.
Customer lifetime value (simple) = AOV × purchase frequency × average customer lifespan in the same time unit.
Enrollment rate = new loyalty sign-ups ÷ total new or eligible customers × 100.
Active rate = members who transacted in the period ÷ total enrolled members × 100.
Redemption rate = points or rewards redeemed ÷ points or rewards issued × 100.
Sales penetration = revenue from members ÷ total revenue × 100.
Incremental revenue = revenue from the test group minus revenue from a matched holdout group, adjusted for group size.
Loyalty ROI = (incremental margin minus program cost) ÷ program cost × 100.

Pick a consistent time window that fits your restaurant or retail cycle, decide whether refunds and returns get netted out of revenue before or after the calculation, and anchor cohorts to either enrollment date or first purchase date, never both in the same report.

Pro Tip: Run every formula on a consistent rolling window first; shorter windows expose seasonality that a full-year average will quietly hide.

Program health vs. business-impact framework: what to measure at each tier and how they connect

Epsilon’s loyalty metrics research lays out a three-tier structure that keeps teams from mistaking activity for impact, and warns specifically against judging a program by vanity numbers like total members or points accumulated.

Tier 1, member behavior: enrollment, active rate, and redemption activity. These move weekly and act as an early warning system, long before revenue trends show up.
Tier 2, program health: retention, repeat purchase rate, and CLV. These confirm whether the behavior changes in Tier 1 are sticking and compounding over months, not just one campaign cycle.
Tier 3, business impact: incremental revenue, incremental margin, and ROI. This is the tier that answers the only question finance actually cares about: did the program make more money than it cost.

The tiers connect in sequence. A healthy Tier 1 without Tier 2 movement means the mechanics attract attention but do not build habits. Healthy Tier 1 and Tier 2 without Tier 3 improvement means you are buying loyalty you would have gotten anyway, which is exactly why incrementality testing sits at the center of a credible measurement program.

Measuring incrementality and attribution: holdouts, A/B, and common attribution traps

Sales penetration and repeat purchase rate can rise while the program itself contributes nothing new, because loyal customers often would have returned anyway. Proving causation, not just correlation, is the job of incrementality testing.

Randomized holdout groups: randomly assign a portion of eligible customers to a control group that receives no loyalty offers, run the test for at least one full purchase cycle, then compare revenue and frequency against the test group.
Difference-in-differences: when a true holdout is not feasible, compare the change in behavior between an exposed group and a similar unexposed group over the same period, which controls for broader trends like seasonality.
Matched cohorts: pair members and non-members on similar spend history, visit frequency, and tenure when randomization is politically or operationally difficult, then compare outcomes between the matched pairs.

Common traps worth watching for: selection bias (your most loyal customers join first, inflating results), channel cannibalization (a loyalty discount simply replaces a purchase that would have happened at full price), and coupon cannibalization (members stack a loyalty reward on top of a public promotion that it never needed to combine with). Mitigate each by keeping holdouts random rather than opt-in, tracking margin alongside revenue, and auditing stacked offers monthly.

Pro Tip: Keep your holdout group small but permanent; a rotating or seasonal holdout resets your incrementality baseline every time it changes.

Benchmarks and target-setting: how to pick KPI targets and interpret performance

EY’s guidance on demonstrating loyalty ROI recommends establishing your own pre-launch baseline and comparing post-implementation performance against it, rather than chasing an industry number that was never built for your cost structure.

Start with a pre-launch baseline for retention, AOV, and purchase frequency, then set percent-change targets tied directly to margin impact rather than round numbers.
Treat outside benchmarks as directional only. CBRE’s hotel loyalty analysis found loyalty members accounted for 52.8% of occupied hotel rooms in 2024, with program costs around $5.46 per occupied room, about 1.6% of revenue, a useful reference point for weighing reward cost against revenue share, but not a target to copy into a different business.
Run a three-step routine: set the business goal (margin growth, retention lift), backsolve the KPI target that would deliver it, then stress-test that target against rising reward costs before locking it in.

Operationalizing measurement: dashboards, segmentation, cadence, and reporting to the C-suite

KPIs only change outcomes when someone sees them often enough to act. Set a cadence that matches the decision being made: daily for operational flags like redemption spikes, weekly for campaign performance, monthly for program health trends, and quarterly for the full ROI conversation with leadership.

Segment reporting by new members, high-tier members, dormant members, and specific campaign cohorts, since blended averages hide the drop-offs that actually need attention.
Build one dashboard as the single source of truth, with metric definitions written down so retention does not mean one thing in marketing and another in finance.
Automate data refresh where possible, and give executives a simple top-line view with the ability to drill into any segment without pulling a separate report.

Pairing dashboard work with field-level tactics tends to pay off faster than either alone; our customer retention tactics for restaurants guide walks through specific moves that move the Tier 2 numbers covered above.

Pro Tip: Give every KPI a single written definition and owner before the first dashboard review, otherwise two teams will argue about whose number is right instead of what to do about it.

Analysis of guest profile data and social login metrics as loyalty indicators

Guest profile data, name, visit history, dietary preferences, and social login details collected at sign-up or checkout, feeds every KPI in this guide. Without a profile tied to a transaction, retention and repeat purchase rate cannot be calculated at the individual level, only estimated in aggregate.

Social login specifically matters because it lowers the friction that usually kills enrollment rate. A guest who joins with one tap through an existing social account is far more likely to complete sign-up than one asked to fill out a form, and the data captured that way, email, name, sometimes birthday, becomes the backbone of segmentation for dormant-member and high-tier drilldowns.

The quality of that profile data also affects how much you can trust your redemption and active rate numbers. A profile with a verified email and a consistent member ID across visits lets you track a guest through the full behavior-to-impact chain; a guest who orders as a one-off guest checkout drops out of every KPI calculation past the point of sale. Statista’s customer-profiling research points to growing demand for first-party profile data precisely because third-party tracking has become less reliable, which makes the profile capture step, not just the loyalty mechanics, part of the measurement system itself.

Author perspective on measurement priorities for loyalty teams

Our honest read: most loyalty dashboards are full of vanity metrics dressed up as strategy. Total members and points issued feel like progress but tell you nothing about margin. Treat loyalty as a line on the P&L, not a marketing badge. Before scaling any change, run a small holdout. Pair Tier 1 engagement numbers with Tier 3 financial checks, never one without the other, or you are measuring motion instead of money.

— Abhi

How MyDigiMenu supports loyalty KPI measurement and program operations

Measuring these KPIs well depends on having clean, connected data in the first place, which is exactly where we built our platform to help. We capture guest profiles through social login, run digital stamp cards, and connect that data to CRM and POS systems to support your retention, repeat purchase, and redemption calculations.

Guest profile capture through social login feeds the member-ID data your Tier 1 KPIs depend on.
Digital stamp cards support loyalty programs and redemption tracking.
CRM and POS integration keeps revenue, visit, and member data aligned for repeat purchase rate and sales penetration.
Targeted campaign tools support targeted marketing campaigns.

Plans start with the StartUp Menu at $39 per month; current prices and plan details are on our pricing page. If you are ready to see how the pieces fit your current data setup, our pricing page is the place to compare plans.

FAQ

How do you measure customer loyalty?

Customer loyalty is measured by combining behavioral KPIs like retention rate and repeat purchase rate with a satisfaction signal like NPS, since NAXION’s research shows NPS predicts outcomes but lags behind actual behavior. The most reliable approach tracks both engagement metrics and financial outcomes like sales penetration, rather than relying on a single score.

What KPIs should you track to measure loyalty program success?

Start with active rate, redemption rate, retention rate, repeat purchase rate, CLV, sales penetration, incremental revenue, and loyalty ROI, organized into the member behavior, program health, and business impact tiers covered above. Zeta Global’s guidance similarly warns against treating enrollment alone as a success metric, since it says nothing about repeat behavior or margin.

What role does a loyalty program play in marketing?

A loyalty program gives marketing a structured way to capture first-party guest data, segment customers by behavior, and run targeted campaigns that are easier to measure than broad advertising. It also creates a direct channel, through sign-up profiles and app or card activity, for re-engaging dormant customers and rewarding high-value ones.

What does KPI mean in branding, and how does it apply to loyalty?

In branding, a KPI is a specific, trackable number used to judge whether a brand initiative, including a loyalty program, is moving the business toward its goals rather than just generating activity. Applied to loyalty, that means prioritizing measurable outcomes like retention and incremental revenue over branding-only metrics like impressions or sign-up volume.