Companies launch loyalty programs to acquire new customers, drive incremental sales, repeat purchases, frequency, or share of wallet. Pick your KPI.

And yet, many fail at what actually matters: building a relationship with their members.

Loyalty is a human relationship, and human relationships are almost always more profitable than the metrics used to justify them.

The problem is simple. Human relationships are hard to measure. Trust, familiarity, and emotional attachment do not sit comfortably in a spreadsheet, whereas a projected five percent sales uplift triggered by the promise of a reward does.

Companies tend to optimize for what is easy to calculate, and easier to get approved internally, rather than for human connection.

For years, Starbucks Rewards had a Gold tier where members received a physical gold card that signaled status and bragging rights every time they pulled it out.

The plastic created friction. Operational complexity in stores. Physical card production and logistics. Training baristas to recognize and treat Gold members differently. So the decision to remove it was straightforward on paper. Way too expensive.

But when Starbucks removed it, complaints were not about losing discounts or a free coffee. They were about losing identity and what made the program feel special.

The gold card worked as a badge of belonging. Being acknowledged by baristas made people feel like regulars. Many said they no longer felt a relationship.

The program continued to change mechanics, focusing on transactions. Which delivered great metrics, for a while. Until the mercenary relationship, sales, started to drop.

Human relationships are better in the long run. They outperform short term incentives because they change behavior, not just transactions. A discount can pull a purchase forward. A relationship changes how a customer feels about choosing you in the first place. It increases price tolerance, forgiveness when things go wrong, and preference when competitors offer something similar or cheaper.

Some airlines have quietly been removing elite recognition rituals. For years, American Airlines made elite status visible and ceremonial. Gate agents called elite groups separately. Boarding announcements explicitly thanked Executive Platinum and Platinum members. Staff often acknowledged status verbally during boarding and inflight.

As the airline pushed for faster boarding and operational efficiency, these rituals faded. Boarding groups were consolidated, announcements became generic, and elite recognition was reduced to a line on a screen.

Nothing major was taken away on paper. Miles, upgrades, and priority lanes still exist.

What disappeared was visibility. Elites were no longer seen or publicly acknowledged. Status became something you knew, not something others noticed.

American Airlines did not cut rewards. They cut recognition. And with it, part of what made elite status feel earned and meaningful. The cost was unmeasurable.

Companies look at costs and think in prices. These are numbers that are easy to work with, tangible, and forecastable, and they naturally become the basis for business decision making.

A classic example of loyalty working exactly as predicted by a spreadsheet is supermarket fuel points, redeemed as cents off per gallon discounts.

For example, Kroger offers fuel points based on grocery spend, which can be redeemed as cents off per gallon at Shell.

The behavior change is measurable. The sales uplift can be modeled before launch and confirmed after, based on assumptions about how many customers will change behavior and how much incremental spend that creates.

Shoppers add items to hit the threshold. Trips consolidate. Baskets grow. Finance gets the result it expected. It works. It is defensible. It is easy to approve.

And it is exactly why loyalty so often optimizes for what is easy to measure.

The challenge appears when this logic comes at the expense of the nuances of human decision making, which is rarely rational or linear. Yes, from a consumer perspective, the Kroger offer makes sense. Who does not want a fuel discount?

But consumers do not always choose what is economically optimal. They choose what is convenient, familiar, emotionally comfortable, or aligned with habit. They may pass on the discount to shop closer to home, avoid an extra stop, or simply because another store feels easier or more familiar.

The spreadsheet predicts behavior. Humans often choose something else.

Multiple retailers have found that a meaningful share of customers do not redeem fuel rewards even when it would be financially rational to do so. Not because they forget, but because redeeming them introduces friction. An extra stop. A different route. A station that feels inconvenient, crowded, or unfamiliar.

From a spreadsheet perspective, driving an extra five minutes to save one dollar is an obvious choice. From a human perspective, it often is not worth it.

Behavioral economists describe this as transaction costs outweighing monetary gain. People consistently overvalue time, mental effort, and routine relative to small financial savings.

Importantly, this does not break the finance model. The program succeeds even when customers do not redeem the full value they earn. The incremental grocery spend already happened. The loyalty cost is contained. The economics work without requiring deeper emotional engagement.

What may erode over time is not the economics, but the meaning of the reward. When a benefit is frequently ignored or bypassed, it risks becoming background noise rather than a source of attachment.

Over time, the financial impact of that detachment will inevitably come. Metrics are rarely irrefutable proof that something is right.

Consumers are not thinking, “I love Kroger because they understand me.” They are thinking, “This is a good deal if it is convenient.” And convenience is not something a spreadsheet can reliably forecast.

To be clear, this is an argument for financial rigor, not against it. Good financial operators understand that rigor includes accounting for human behavior, even when its impact arrives later.