Loyalty Program Do's and Don'ts: A Practical Checklist

Loyalty program do's and don'ts are the practical rules of thumb that guide a program from its first objective through launch and day-to-day management: what to reward, how to make the program easy to use, how to keep it relevant, how to measure it, and how to treat members' data and trust. This guide sets out ten of them as paired do's and don'ts. None of these are absolute laws; context matters, and there are exceptions. As rules of thumb, though, they keep most programs out of most trouble.

The stakes are real. McKinsey found in 2021 that around two-thirds of established loyalty programs fail to deliver value, with many actually eroding value. That figure does not say why each program fell short, and not every cause is in a program team's hands. This guide focuses on the ones that are: what the program rewards, how hard it is to use, how it is measured, and how honestly its economics are tracked.

Here are the ten pairs at a glance. The sections that follow explain each one, how to apply it, and where the exceptions lie.

Do start with a clear objective tied to business value. Don't launch a program without knowing what it is for. Every later decision, from the earn rate to the reward catalog to the reporting, depends on what the program is meant to change: purchase frequency, retention, basket size, cross-category buying, or, in a channel program, sell-through and partner engagement. Matching a competitor's program can be a legitimate objective, for example to stop members defecting to a rival's scheme, but it then needs to be written down and measured like any other. A program launched only because competitors have one tends to copy their mechanics and inherit none of their reasons. Write the objective down as a measurable outcome, such as "raise 12-month retention among second-year customers", and check each design choice against it.

Do reward the behaviors the program actually needs. Don't reward spend alone, or the wrong behavior. A program produces more of whatever it pays for. Reward only transactions and it may subsidize purchases members would have made anyway; reward a rich sign-up bonus and it can attract deal-seekers who claim it and leave. Decide which behaviors drive the objective, such as trying a new category, completing a profile, referring a friend, or engaging with content, and build the earning rules around them. In channel programs, the equivalent trap is rewarding sell-in, which can reward distributors for loading inventory at the end of a quarter rather than for selling it on.

Metrolink, Southern California's commuter rail, applies the same idea in transit: its SoCal Explorer program on BLOYL™ awards points for every mile traveled and rewards both transactional and non-transactional rider behaviors. Among enrolled riders it recorded a 60% active engagement rate, and members showed a 15% increase in average monthly transactions (metrics disclosed by Brandmovers). The case page does not report a holdout comparison, so neither figure isolates what the program added.

Do make joining and using the program effortless. Don't bury it behind needless friction. Each extra field, password rule, or step between interest and a first reward is another point where people can drop out. E-commerce research shows the cost of forced steps: in Baymard Institute's checkout research, updated in September 2025, 18% of shoppers who abandoned a checkout, setting aside those who were just browsing, gave "the site wanted me to create an account" as a reason. Loyalty enrollment is a different moment from checkout, but the lesson is a reasonable one to carry over: ask only for what the first interaction needs, make the first reward reachable early and easy to redeem, and collect the rest of the profile over time as members see value.

There is an honest exception. A small amount of friction can be useful, because it filters for members who genuinely intend to take part, and some programs ask for more information up front because they need it to deliver the benefit. The test is whether each step earns its place.

Do start simple and improve over time. Don't over-engineer the program at launch. Multiple currencies, many tiers, and category-specific multipliers each look reasonable on their own; stacked together, they leave members unable to work out what a purchase is worth. Launch with rules a member can grasp in seconds, then add complexity only where data shows it pays. Simplicity has a limit, too: simplify too far and the program becomes indistinguishable from every other points scheme. Before launch, confirm that the point-of-sale, e-commerce, and CRM systems can capture every behavior the program rewards; a behavior the program cannot track cannot be rewarded reliably or measured. The guide to the five pitfalls of loyalty program design covers that trade-off.

Do personalize using the data members share. Don't send everyone the same generic offers. A loyalty program holds each member's purchase, earning, and redemption history, which puts it in an unusually strong position to be relevant. Consumers have come to expect it: in McKinsey's 2021 personalization research, 71% said they expect companies to deliver personalized interactions, and 76% said they get frustrated when that does not happen. The same research put the revenue gap at 40% more from personalization activities for the companies that do it best, compared with average players, which is a comparison between companies rather than proof that personalization alone produced the difference. Personalization has limits as well: use data members can see the reason for, honor their preferences, and stop short of messages that feel intrusive.

Do build a relationship beyond points. Don't compete only on who pays the richest reward. A program that offers nothing but discounts competes on the one axis every rival can match. Recognition, service, access, community, and useful content give members reasons to stay that a competitor cannot simply outbid. Gallup reported in 2014 that a fully engaged customer represents an average 23% premium in share of wallet, profitability, revenue, and relationship growth compared with the average customer. The data is dated, and it measures engagement rather than any program feature; it also cannot show whether engagement drives value or valuable customers simply report more engagement. What it does show is that the customers who feel most attached are worth more than average, which is the reason to look beyond price.

In B2B, a leading Canadian regional distributor built its Culture Club program on BENGAGED™ around community and brand advocacy, deliberately avoiding a transaction-only design. Customers who enrolled grew their sales by 25% on average, against 5% for those who did not (metrics disclosed by Brandmovers). Members chose to join, so part of that gap may come from who enrolled rather than from the program itself.

The exception: not every customer wants a deep relationship. Some members join for the transactional benefit and nothing more, and a program should serve them well rather than push them into engagement they do not want.

Do keep the program fresh and stay in touch. Don't set it and forget it. Members lose interest in a program that never changes, and they forget one they never hear from. Plan a refresh calendar for rewards, challenges, and offers, and set up messages tied to what members do: a welcome after enrollment, a nudge near the next reward, a reminder before points expire, and a re-engagement message when activity drops.

Staying in touch is not the same as sending more. Each unneeded message uses up some of a member's patience and makes an unsubscribe more likely, so set frequency caps and let members choose how often they hear from the program. When a change reduces members' existing value, such as a new earn rate or a reward being removed, give advance notice and explain the reason. The guide to building a customer loyalty communication strategy covers what to send, through which channels, and when.

A loyalty program does not need to be perfect to succeed, and avoiding the common, avoidable mistakes removes many of the ways it can fail. Read together, the ten pairs come down to four ideas: reward the relationship the program actually wants, make that easy for the member, prove the program's value honestly, and respect the person on the other end. Most of the do's and don'ts are specific applications of those four.

They are rules of thumb, not an exhaustive list. Some friction is useful, some members only want the transaction, and every program's own data should refine them. The guide to understanding the value and ROI of customer loyalty programs covers how to make the financial case once the basics are in place.

The case, in numbers