Why Age and Income Don't Define Brand Loyalty Anymore

Demographic segmentation groups customers by attributes such as age, income or location. In loyalty programs it helps with reach and reporting, but on its own it is a weak guide to why a member stays, because loyalty depends on attitudes and behavior that vary widely within any age or income band.

It is tempting to start a loyalty plan from age bands and income brackets, because those categories are easy to buy and report, and generational data does show real average differences. But an average cannot tell a program how any one member will behave. What has changed is the data: loyalty programs capture transaction and engagement data, so age and income no longer have to stand in for behavior. This guide covers what generational data actually shows, why age and income are weak proxies on their own, what to segment on instead, how to handle consent when personalizing and how to test whether a new segmentation works better than the old one. For generation-specific program design, see the guide to Millennial vs Gen Z loyalty.

Key Takeaways

What does generational data actually show about loyalty?

Generations differ on average: younger members are more willing to share data for personalization and more drawn to causes, community and digital features. Averages still hide wide variation.

Deloitte's 2025 survey of 5,564 US adult loyalty program members found that "89% of Gen Z and 87% of millennials surveyed are willing to share personal information for more tailored offers or experiences, compared to 78% of Gen X and 64% of baby boomers." Asked about hyper-personalized loyalty settings, 62% of Gen Z and 64% of millennials said they would opt in, versus 55% of Gen X and 33% of boomers, and 51% of Gen Z and 53% of millennials said they would spend more with a personalized experience, versus 38% of Gen X and 19% of boomers.

Deloitte also reports that Gen Z and millennials place more importance on "contributing to missions and causes, opportunities to participate in community member events, and efficient and enjoyable digital experiences" (Deloitte). Respondents were existing loyalty members in seven consumer industry groups, key questions were asked about each respondent's most-preferred program, and the answers are stated intentions, not observed purchases. Deloitte notes the results "should not be generalized across all loyalty memberships." A single survey also cannot show whether these gaps reflect generation or life stage, or whether they are changing over time. Age therefore matters for defaults, such as how much personalization to offer and how to ask for data, but a third of boomers still said they would opt in to hyper-personalized settings, and 38% of Gen Z did not say they would.

Why are age and income weak proxies for loyalty?

Loyalty combines how members feel about a brand with whether they keep buying. Age and income measure neither directly, and loyalty varies widely within any age or income group.

Dick and Basu describe customer loyalty as "the strength of the relationship between an individual's relative attitude and repeat patronage" (Journal of the Academy of Marketing Science, 1994). Demographics describe who a customer is, not how they feel about the brand or how they behave. As an illustration, not a benchmark: two 35-year-olds with similar incomes might be loyal for opposite reasons, one because a brand saves time on a busy weekly routine and the other because its sourcing matches values they care about. A program aimed at "35 to 44, middle income" would give both the same offers. Income shapes what customers can afford and which rewards feel worthwhile, so it can matter for price points, paid tiers and reward mix. But income measures ability to spend, not attitude or repeat patronage. This guide cites no data showing income predicts loyalty on its own, so treat claims that one income group is "more loyal" as hypotheses to test in your own data.

What should you segment on instead?

Combine behavior from transaction and program data with stated needs and values, and keep demographics for reach, reporting and legal requirements such as age limits.

As an illustration, not a benchmark: a grocery program might start with four behavioral segments: weekly members who buy across several categories, weekly members who buy in one category, members whose gap between visits has doubled and members in their first 90 days. A two-question preference survey (what matters most: price, convenience, quality or causes) then splits each behavioral segment by stated need, so rewards and messages can differ within a segment that buys the same way. New members have little behavioral history, so start them on stated preferences and sensible defaults, then shift to behavior as data builds. The guide to dynamic segmentation covers how segments can update as behavior changes.

How do behavior and stated needs work together?

Behavior shows what members do; stated needs and values help explain why. Together they help separate members who are loyal from those repeating out of habit or convenience.

A member who buys every week may be loyal or may simply live nearby; a member who buys less often may still prefer the brand strongly. Purchase data alone cannot tell them apart, which is why short preference questions, asked when members get something in return, add context behavior cannot. Use behavior to set earn rules, timing and win-back triggers, and use stated needs to choose rewards and messages. BLOYL™, Brandmovers' loyalty platform, lets marketing teams configure earning and redemption rules by customer segment and behavioral action without engineering work, and includes A/B testing against a control group and predictive churn analytics. The guide to the psychology of customer loyalty covers the habitual-versus-true loyalty distinction in more depth.

How do values and experience shape loyalty across segments?

Values and consistent experience matter across ages, but how much varies by person. Back any values message with visible action, and test which experiences matter to which members.

Deloitte's finding that younger members give more weight to causes and community is an average, not a rule; an average gap does not mean older members do not care about the same things, or that younger members cannot care mainly about price and convenience. Values-based messaging also carries a risk if a brand's actions do not match its claims, so tie each values message to something members can check, such as a named partner, a donation total or a sourcing standard. Consistent experience, such as rewards that arrive when promised and service that solves problems, is less dependent on values and more on operations, and it can be measured directly through complaints, redemption completion and retention.

How should you handle consent when personalizing?

Personalizing on behavioral data works best when members understand and agree to it. Offer clear choices, especially for members who are less willing to share.

Deloitte's data shows that willingness to share personal information for tailored offers falls with age, from 89% of Gen Z to 64% of boomers, so a single data request will suit some members and put off others. Explain what members get for sharing, let them choose the level of personalization and make it easy to change their mind. In California, a business covered by the CCPA that offers financial incentives tied to personal information must give notice and obtain the consumer's prior opt-in consent, which "may be revoked by the consumer at any time" (Cal. Civ. Code 1798.125). This is general information, not legal advice.

How do you test whether a new segmentation works better?

Build the new segments, give some members in each a different treatment and compare them with a holdout. Keep the demographic version as a baseline to beat.

A segmentation is better only if it changes decisions that improve results. Pick an outcome, such as retention or response to offers, and compare treatments designed for the new segments against both a holdout and treatments designed for demographic groups. Set the outcome, the holdout share and the test length before launch, size each group so it can detect a meaningful difference, run the test across at least one full purchase cycle and judge results on the pre-set outcome, not whichever metric looks best afterward. Pair behavioral results with attitude measures, such as stated preference for the brand, so the test can separate loyalty from habit. The guide to emotional loyalty covers how to measure attitudes.

Conclusion

Age and income still say something about members on average, and generational data shows real differences in how people want loyalty programs to work. But an average cannot predict any one member, and loyalty depends on attitude and behavior that vary widely within every group. Segment on what members do and what they say they need, use demographics for defaults and reporting, handle data with clear consent and test any new segmentation against the old one before rebuilding around it.

Sources

• Deloitte Insights, "Reshaping loyalty programs in an era of value seeking" (2025 Deloitte Consumer Loyalty Program Survey)
• Alan S. Dick and Kunal Basu, "Customer Loyalty: Toward an Integrated Conceptual Framework," Journal of the Academy of Marketing Science 22(2), 1994
• California Civil Code section 1798.125