Launching was the easy part: What comes next?
Every loyalty program has one honeymoon moment: the week after launch, when the enrolment dashboard is climbing and the whole team feels like the program is already working. I get why. It's the easiest metric to move, and for the first few weeks it's the only one you have. That's also exactly why it's the easiest number to misread.
A launch spike tells you marketing worked. It tells you nothing about whether the program works, and mixing those two up is where loyalty program optimisation actually starts, or stalls, for most teams I work with.
What's the onboarding gap?
The pattern I look for immediately is high enrolment plus low engagement. That combination is almost always an onboarding problem, not an acquisition problem. You brought people in, and then didn't give them a reason to act in the first 30 days. If members grow but activity doesn't, the funnel is front-loaded, and every additional enrolment just makes the denominator worse.
So the honest post-launch question isn't how many people signed up. It's how many did something. Answering that well means watching more than one number, which is where most teams' measurement actually needs to start.
What are the four stages that actually matter?
I use the same four-stage funnel with every client: enrol, engage, retain, ROI. Each stage answers a different question, and each one has a benchmark band that turns a status update into an actual decision.
Enrol comes first, and it only answers one thing: are you reaching the right companies and members? I watch new companies enrolled, new members, eligible reach, and enrolment velocity here. A healthy number looks like positive year-over-year member growth, with enrolment reaching somewhere between 30 and 50 percent of the eligible audience.
From there it's engage, whether members are actually becoming active instead of just signing up. Active members, training completion, claim participation, portal logins, redemptions, that's the watch list. I want to see an above 40 percent claim rate and 35 to 55 percent active monthly participation.
Retain is where it gets more interesting: are the behaviours actually repeating and sticking? Average earn value per active participant, period-over-period retention, recency distribution, cross-sell. A healthy retain stage runs 55 to 75 percent of next-period retention, with earn value climbing and a stable recency mix.
Then ROI, which answers the question everyone actually cares about: is the program influencing revenue efficiently? Revenue influenced, the earn-to-revenue ratio, eligible revenue, member versus non-member revenue, incremental revenue. The floor is 1x earn-to-revenue. Best-in-class programs run 4:1 to 8:1, with both influenced and incremental revenue positive.
The discipline is reading these in order. A weak ROI number is usually a retention symptom, and a weak retention number is usually an engagement symptom, which is exactly why fixing the last stage first is the most common, and most expensive, mistake I see. Below 1x earn-to-revenue, a program isn't sustainable, you're buying revenue you already had. The benchmarks aren't there to grade the program. They're there to tell you which stage to go open up when a number falls outside the band.
The bands tell you where to look. What's actually happening inside them takes a sharper kind of reading.
How do you read the signals before they become problems?
I read metrics in pairs, because a single number rarely tells you anything on its own. Three combinations do most of the diagnostic work:
High enrol and low engage points to an onboarding issue; tighten targeting and fix within the first 30 days.
Strong engage and weak retain means behaviour isn't sticking, so add repeat triggers, tiers, and better communications.
Strong retain and weak ROI means the program economics are out of balance; rebalance earn-to-reward and focus on high-value behaviours.
Two more worth watching for:
Logins without claims usually means claim friction or an unclear value exchange—people are browsing, not converting.
Low redemption despite real activity points at the catalogue or the reward thresholds, not at the members.
The one almost nobody watches is engagement recency distribution. It tells you who's about to churn a quarter before your retention rate does. When participants start sliding from the 0–30 day bucket into 31–60 and 61–90, that's your early warning, and it's the window where a win-back campaign is still cheap.
Diagnosis only matters if it leads somewhere. Match the lever to the leak, and resist the urge to pull all of them at once. Enrolment below target calls for a recruitment campaign, sharpened targeting, and training support. Weak activation calls for an onboarding journey, a first-action bonus, and segment optimisation. A weak repeat rate calls for a reminder cadence and a win-back campaign before you touch the onboarding flow itself. A soft E:R ratio calls for offer and segment optimisation over adding volume anywhere else in the funnel.
More recruitment is the right answer to exactly one problem on that list, and it isn't the one most teams are having. That's the mistake I see most often: treating every problem like a marketing volume problem, when the pattern above says otherwise three times out of four.
Diagnosing well only gets a team halfway there. The rest comes down to a rhythm that surfaces these problems before they compound.
How do you build a rhythm around the numbers?
Separate monitoring from decision-making. Weekly WIPs are for monitoring. QBRs are for decisions. If a QBR gets spent reviewing numbers everyone already saw last week, it's the one meeting where anyone was actually empowered to change something, and it just got used to restate the obvious.
I run these in 30 minutes, no longer: the headline of what changed, funnel performance, which segments are actually driving the result, the financial guardrails, and a decision log. It isn't finished until it produces one insight, one decision, one KPI target, and one owner.
Get that rhythm right and the numbers start compounding on each other instead of resetting every quarter. Get it wrong, and the last mistake mostly takes care of itself. Confusing activity with impact is the biggest one I see after launch. Enrolment gets reported because it's flattering and it moves early, but a program's job is to change behaviour in a way that produces revenue that wouldn't have happened otherwise. That's why member versus non-member revenue and incremental revenue matter more than influenced revenue. Influenced revenue shows reach. Incremental revenue is the only number that proves the program created real growth instead of just tracking sales you'd have made anyway.
What actually comes next?
Launch is the beginning of the measurement problem, not the end of it. Pick the four stages, agree on a benchmark band for each, and run the same three questions every quarter: what changed, why did it change, and what do we do next. Then act on the biggest leak in the funnel, not the loudest number on the dashboard.
That's what loyalty program optimisation actually looks like after go-live. Not a single launch metric, but a discipline you repeat until the whole funnel starts building its own momentum.
If your program's past launch and the numbers feel murky, talk to our team and we'll help you figure out which stage actually needs the attention.
