Are You Measuring Gen Z Loyalty in the Wrong Place?

A recent study of observed US spending behaviour offers a useful challenge to the way brands commonly discuss younger customers and loyalty.

Attain and dentsu’s analysis of purchase data from September 2025 to August 2026 found that Gen Z spent around 25% less annually than the average US shopper, while purchasing from 6% more merchants. The group was also more likely to say brand names mattered, but less likely to say it usually bought from the same brands.

That combination is more interesting than the familiar claim that Gen Z is disloyal. It suggests that a conventional retention dashboard can give brands the wrong diagnosis. A lower repeat rate may not simply reflect dissatisfaction or weak CRM. It may reflect a customer who has a limited budget, a wider set of competing merchants and a different mix of occasions on which to spend.

The practical question is not whether a brand can make every Gen Z customer faithful. It is whether it knows which moment in the customer’s life it is trying to own, and whether its evidence captures the alternatives that compete for that moment.

Key takeaways

• Define competition by occasion as well as category. A restaurant, fitness subscription, gaming purchase or rideshare journey may be competing for the same discretionary spend.

• Separate trial, repeat and wider wallet share. One retailer’s customer file cannot show whether a lost purchase went to a direct rival, another category or no purchase at all.

• Treat loyalty as a portfolio of behaviours. Regular purchase, recommendation, direct search, willingness to pay full price and advocacy do not always move together.

• Use younger customers’ switching as a research prompt. Find the moment at which a habitual choice becomes open to reconsideration, rather than assuming an offer alone caused the change.

Loyalty measurement becomes less useful when it records the transaction but misses the occasion that prompted it.

A smaller wallet can produce noisier signals

The Attain and dentsu findings should not be transplanted uncritically into the UK. They are US data, based on the provider’s purchase-data panel and indexed against an average US shopper. Income, transport, housing and retail structures all shape what a young adult’s wallet looks like.

But the measurement problem travels well.

When spending is dispersed, a brand sees only a partial version of customer behaviour. Its own data may show a lapsed customer, a less responsive subscriber or a falling frequency segment. It cannot, by itself, reveal whether that customer has switched to a competitor, substituted an adjacent experience, deferred the purchase or reduced discretionary spending altogether.

This is particularly important in categories where Gen Z’s spending moves between goods and experiences. A beauty brand is not only competing with beauty brands. A casual dining operator is not only competing with other restaurants. The relevant set may change by payday, social plans, season, location and the customer’s immediate need for convenience, identity or reward.

That does not make first-party data less valuable. It makes it incomplete. Brands need to combine it with category behaviour, customer interviews and, where feasible, external transaction or panel data. The goal is not a grand unified picture of every purchase. It is a credible view of what customers are choosing instead.

Brand preference is not the same as repeat purchase

The study’s apparent contradiction — higher interest in brand names alongside lower stated repeat-brand preference — is a reminder that brand strength has several layers.

A customer may value a recognisable brand because it reduces perceived risk, signals taste or makes selection easier. Yet that same customer may still experiment when a new product, outlet, creator recommendation or price point appears. In this setting, awareness and affinity can create a place on the shortlist without securing the next transaction.

Other recent US research points in the same direction, although with a different method. Attentive’s survey of 600 shoppers found that many had tried a new brand recently, while most said they regularly shopped with only a small group of brands. Its respondents put product quality, value and personal relevance ahead of membership of a rewards scheme when explaining why they sought out a brand by name.

The implication is not that loyalty programmes have failed. They can reduce friction, recognise repeat customers and create reasons to return. But points and personalised offers cannot substitute for a clear reason to choose the brand at a specific occasion.

For insight teams, this means avoiding the temptation to collapse every loyalty question into one score. Ask separately: which brands do customers consider automatically; which do they purchase repeatedly; which do they recommend; and which do they return to without a prompt? The gaps between those answers are often where the useful commercial work begins.

Research the switching moment

A useful next step is to design research around the decision that opens the door to switching.

That may be a first move to a new neighbourhood, a change in work patterns, an invitation from friends, a new dietary preference or a desire to spend less time comparing options. The relevant trigger will differ sharply between categories. CivicScience’s recent analysis of US loyalty, for instance, argues that the attitudes associated with loyalty vary between areas such as automotive, clothing, beauty and fast food. Demographics alone are a blunt guide.

This calls for a mixed approach. Transaction data can establish where and how often behaviour changes. Qualitative work can reveal the jobs, trade-offs and language behind it. Surveys can then test whether the pattern is broad enough to influence positioning, proposition or communications.

Brands should also resist treating a more fragmented journey as entirely digital. eMarketer’s October path-to-purchase research argues that physical stores remain central across an increasingly digital shopping process. For many customers, a social post, search result, AI answer, friend’s recommendation and store visit may all contribute different forms of reassurance.

The winning measurement framework will therefore be deliberately narrower than the promise of total personalisation. It should identify the few occasions that matter most, the alternatives present at those moments and the evidence that a customer has genuinely moved from interest to preference.

Gen Z’s wider merchant behaviour is not evidence that brands should give up on retention. It is evidence that retaining a customer starts with a more realistic account of what the brand is competing against.