What loyalty programs get wrong about consumer behavior

Myth No. 1: The 80/20 Rule

• The assumption: A small number of customers (~20%) drive most sales and value (~80%), justifying a focus on a small number of high-value customers.
• What the data shows: Analyzing data from many client loyalty programs shows that instead of 80/20, the real ratio is closer to 50/20. The top 20% of customers generate about half of loyalty program-related revenue in any given time period. Additionally, this 50/20 ratio is found to be true not only within the loyalty program, but also more broadly across a brand’s entire customer base. “This means that half of a business’s revenue comes from its ‘lighter’ users, O’Connell said, “and yet less frequent customers are too often ignored when crafting loyalty programs.”
• Example: Over a 14-year period for a multi-billion-dollar specialty retailer, loyalty program sales grew over time, but the percentage of sales from the top 20% of customers was remarkably consistent at around 55%, something we observe repeatedly across various loyalty programs. Sales success comes from appealing to all customers, not the most loyal, especially when we consider Myth No. 2.

• Marketing strategy takeaway: If half the sales in most customer bases come from low frequency customers and the distribution of behavior doesn’t change over time, it’s critical for marketing strategies to engage the whole customer base, not just the top customers. Growing the overall customer base will grow all frequency levels. “It’s nearly impossible to get more volume from the top spending customers. Growth comes from a balanced approach to increasing the brand penetration with the entire customer base,” O’Connell said.

Myth No. 2: Top customers are a stable cohort

• The assumption: A brand’s best customers are a definable group that stays relatively consistent over time. Brand stability is built on serving a core group of intensely loyal customers.
• What the data shows: The actual composition of the top 20% of customers is constantly changing over time and is not a group of similar customers.
• Example: When we look at top customer groups over time, whether using decile analysis or RFM analysis (recency, frequency, monetary), we find an unexpected paradox. The top group has a consistent impact on sales, but the members in the top group are always changing. For a restaurant client we worked with, 70% of the top 20% of customers in their loyalty program were different from one year to the next. For a national big-box retailer using RFM analysis,  more than 50% of their top customer group was different a year later.
• Marketing strategy takeaway: Frequency-based segmentation schemes aren’t reliable predictors of future frequency. How frequent a group of customers is during a particular time frame is a valuable business metric, but it doesn’t provide insights into their needs, buying patterns, or values, which is what is needed to create relevant personalized messaging.

Myth No. 3: Frequency equals loyalty

• The assumption: Frequent customers are the most loyal customers. Loyalty programs work best by targeting this segment to create consistent, reliable purchase frequency.
• What the data shows: Customer behavior tends to be “randomly frequent,” especially among the current higher-frequency customers. Customer demand is naturally uneven (driven by needs) and frequency regresses to the mean over time. Today’s top customers will likely not be tomorrow’s top customers and vice versa.
• Example: The loyalty data we’ve worked with in restaurants and retail shows that there is a very high volatility rate in frequency and purchase behavior appears “erratic.” This is the case because needs drive purchase occasions more than any incentive, including discounts, and needs for most product categories are not constant.
• Marketing strategy takeaway: “Consumers have changing needs, and it is critical to meet them where they are to win their business,” O’Connell said. Segmenting customers based on behavioral patterns leads to opportunities to personalize messaging based on the needs and preferences of behavioral segments. Insight-driven personalization strategies have proven to drive better customer engagement as measured by email open and click-through rates.

Myth No. 4: Rewards can drive behavior changes

• The assumption: Incentives and rewards can change customer behavior.
• What the data shows: Customer purchase behavior is inconsistent, and at least half of customer rewards are never used in nearly every loyalty program. It is often difficult to measure the incrementality of loyalty purchases.
• Example: Redemption rates for a national big-box retailer and a fast casual restaurant chain we worked with both hovered near only 25%. Brands consistently see less than 50% of rewards redeemed. The only exceptions are where rewards are presented at the moment of purchase, allowing the customer to take a real-time discount.
• Marketing strategy takeaway: “Consumers won’t buy products they don’t need, even with a reward or incentive,” O’Connell said. In many cases, rewards are so focused on driving incremental behavior that they don’t align with customer needs and, as a result, the customer doesn’t value them. In other cases, the reward value is not compelling enough to create value for customers. Generally, reward structures that offer more choice and that can engage multiple frequency levels have a better chance to align with customer needs and drive more engagement. Even better, personalized messaging strategies have been proven to drive better engagement and can actually reduce the need for rewards while resulting in increased brand affinity and sales.

Moving from myths to engagement and success

The key to success is in a strategic mindset shift from trying to incent more demand to capturing the demand that exists. The most successful loyalty approaches engage customers based on their needs and preferences. Building the right segmentation models — models that are data-driven and seek to understand customer behavioral patterns — do exactly that.

To build behavioral segmentations, brands need the right tools and skills. Most loyalty tech platforms don’t have the analytical capabilities to do behavioral analytics. Once behavioral or preference-based segmentation schemes are defined, marketing teams need to then develop the strategies and leverage technology to actively segment and personalize messaging to generate higher levels of brand engagement, stay top of mind, and, ultimately, capture demand when customers are in the market.

To this end, getting the marketing tech stack aligned with a more analytics-based approach is on the minds of CMOs and other C-suite leaders. Grant Thornton’s 2025 Digital Transformation Survey listed “CRM and customer interface” as the second most frequent choice for digital-related spending priorities. These systems are table stakes for effective loyalty program personalization and execution.