Why Loyalty Programs Fail: The Mercenary Trap

Most loyalty programs don’t build loyalty. They build mercenaries. People who stay as long as your deal is better, and leave the moment a competitor offers 10% more. If you’ve been wondering why loyalty programs fail, why customers accumulate your points and still defect, the answer is hiding in a behavioral science principle called the Over-Justification Effect. It explains why small extrinsic rewards are often worse than no rewards at all. Once you see the trap, you can’t unsee it.

Speed Run Notes

• Small extrinsic rewards trigger the Over-Justification Effect: they switch customers from “I enjoy this brand” into “I’m being paid badly for this engagement.”
• Most loyalty programs run a predictable 5-phase Death Spiral: honeymoon → reality math → reframe → trap → defection.
• The Mercenary Test diagnostic: would your users leave for 10% better rewards elsewhere? If yes, you have mercenaries, not loyalists.
• Escape routes: CD1 Epic Meaning reframe, status that unlocks real capability, right-brain CD5 + CD7 engagement, cross-product value exchange, and Earned Lunch + Free Lunch blends.

The $2 That Ruins Everything

Imagine you spent all day helping a friend move furniture. You hauled boxes up three flights of stairs, disassembled a bed frame, wrestled a couch through a narrow doorway. You did it because you’re a good friend. You didn’t expect anything.

At the end of the day, your friend pulls out their wallet and hands you $2.

How do you feel?

Insulted, right? Not grateful. Insulted. Your brain doesn’t compute “friendship plus $2 equals even better friendship.” It flips into transaction mode and evaluates the deal: “You think my entire Saturday is worth two dollars?”

Here’s what’s strange. If your friend had offered you nothing at all, you would have driven home happy. The $2 didn’t add to your goodwill. It contaminated it. It replaced a story about helping a friend with a story about being paid terribly.

This is the Over-Justification Effect. And it’s silently destroying most loyalty programs, reward systems, and engagement campaigns in existence today.

I’ve spent over two decades studying what motivates human behavior. I’ve applied the Octalysis Framework to products at companies like Microsoft, Porsche, LEGO, and MrBeast’s operations. The single most consistent mistake I see, showing up everywhere from enterprise software to mobile gaming to retail loyalty cards, is what I call the Mercenary Program Trap.

The trap works like this: you try to reward people for doing something, and you end up with people who only do it for the reward. The moment a competitor offers 10% more, they’re gone. You haven’t built loyalty. You’ve built a mercenary army that will defect at the first better offer.

The Over-Justification Effect is one of the most counterintuitive findings in behavioral science. The original research by Lepper, Greene, and Nisbett (1973) showed that children who were rewarded for drawing (something they already loved) lost interest in drawing once the rewards were removed. The external reward had contaminated their internal motivation.

Decades later, the pattern holds across every domain I’ve studied. A church volunteer who loves serving his congregation gets hired as paid staff doing the exact same work. Within months, he’s complaining about his low salary. He told me directly: “I know I shouldn’t feel this way, but I can’t help it.” The extrinsic frame replaced the intrinsic one, and the replacement was worse.

The reason is mechanical. The human brain cannot simultaneously hold both frames. Once you introduce a transaction, the brain has to evaluate the deal. And most reward programs offer terrible deals when measured carefully: you earn 500 points after spending $200, and those 500 points might be worth $1.50. You’ve been paid 0.75% for your loyalty.

That’s not a reward. That’s an insult disguised as a benefit.

Without the points program, customers might have actually enjoyed your product or service. After enrolling, they’ve anchored to a transaction. The transaction is bad. They feel used.

How the Mercenary Program Trap Works

In the Octalysis Framework, I classify motivation into 8 Core Drives. The fundamental axis that matters for loyalty program design is Left Brain vs. Right Brain.

Left-brain Core Drives (CD2: Development and Accomplishment and CD4: Ownership and Possession) are extrinsic, goal-oriented, and analytical. They engage when users feel they’re accumulating, progressing, or winning.

Right-brain Core Drives (CD3: Empowerment of Creativity, CD5: Social Influence and Relatedness, CD7: Unpredictability and Curiosity) are intrinsic, experience-oriented, and relational. They engage when users feel creative, connected, or surprised.

Loyalty programs almost universally operate on the left side: accumulate points, reach a tier, redeem for reward. The problem is that left-brain engagement has a ceiling. Once users calculate the math and find it unfavorable, the motivation collapses completely. And the right-brain engagement they might have had before the program was introduced is gone too.

The real danger isn’t just that the program fails. It’s that the program damages what was already working.

Think of it as trading gold for copper. Before your program, users had a real relationship with your brand. After your program, they have a transaction. A bad one. You made an exchange you didn’t know you were making.

The Loyalty Program Death Spiral

This is the predictable progression I see at companies again and again:

Phase 1: The Honeymoon. User does an activity they already enjoy and discovers they earn points for it. “Oh, this is great. I get rewarded for what I already do.” They’re excited. Engagement spikes.

Phase 2: Reality Sets In. After weeks or months of earning points, the user looks at what they’ve accumulated. Maybe $2-3 in redemption value. They start tracking their effort against the reward. The math is brutal.

Phase 3: The Reframe. The user’s mental model shifts from “I enjoy this” to “I’m being paid 50 cents a month for this.” Once the transactional frame is active, it’s sticky. The user can no longer experience the activity as purely enjoyable. There’s a pay rate attached, and it’s embarrassingly low.

Phase 4: The Trap Closes. This is the part nobody talks about. Without the program, the user got no reward but enjoyed the activity. Now they’re in a state worse than baseline. They’ve been told their behavior is worth something, and the number is insulting. Even quitting the program doesn’t restore the original enjoyment because the transactional frame is now hardwired into how they experience the product.

Phase 5: Defection. The user leaves when a competitor offers slightly better terms. And they feel entirely justified doing so, because this was always a transaction to them, and they’re getting a better deal elsewhere.

You spent resources building a program that actively made your best customers more likely to leave.

The World of Warcraft Cautionary Tale (A Three-Act Warning)

Few case studies illustrate the Mercenary Program Trap as completely as World of Warcraft’s trajectory. Not because it’s about a loyalty program specifically, but because it shows exactly what happens when a thriving engagement system gets contaminated by extrinsic monetization.

Act 1: The Pure Model

WoW launched with a clean economic model: $60 for the box, $10-15/month subscription. Everything in the game, every mount, every piece of gear, every achievement, was earned through player effort. The game made over a billion dollars. Players felt their accomplishments were real. Peak subscriptions hit 12 million players worldwide in 2010.

The key word there is “felt.” Players weren’t just accumulating things. They were accumulating proof of effort. Every rare mount said “I spent 200 hours earning this.” That’s Core Drive 2: Development and Accomplishment at its purest — achievement that carries earned meaning.

Act 2: Freemium Creep

After Activision acquired Blizzard, WoW began adding purchasable shortcuts: exclusive mounts, skins, and progress boosts for real money. The prices weren’t even outrageous. But what happened next was predictable from an Octalysis perspective.

Players who had grinded 300 hours for a rare mount watched someone buy the equivalent for $150 without effort. The earned meaning of their achievement evaporated. Their 300 hours didn’t prove mastery anymore. It proved they hadn’t paid. “Everything becomes meaningless and stupid” was how multiple content creators described it.

Subscriptions collapsed from 11 million to barely a million. The engagement system wasn’t just weakened. It was poisoned.

Act 3: The Final Betrayal

Blizzard responded by releasing WoW Classic, the original 2004 experience, preserved exactly. The player response was massive. The trust was briefly restored.

Then they added pay-to-progress options to Classic “to make it more accessible.”

Content creators who’d built audiences of hundreds of thousands specifically around WoW Classic walked away publicly. Some had been making WoW content for 10+ years. The trust was irreparably broken, not because players were irrational, but because they’d been promised something specific and had it taken away again.

The root cause, in every case: decision-making shifted from “we love the game” to “if we spend $50K on a feature, it must generate $50K back quickly for the CFO.” That’s Black Hat Core Drive 8: Loss and Avoidance driving product decisions, producing short-term revenue at the cost of long-term engagement.

A related pattern shows up in mobile gaming. I once interviewed a player who had spent $30,000 on a single mobile game in one year. It started with a small purchase, then escalated into a spending war with another player on the same server. His logic: “If I’ve spent $30,000 and I’m #2, that would be so stupid. Just spend another $500 to get there.” CD8 creating a self-reinforcing spiral where sunk cost makes each additional dollar feel more rational, not less.

The Mercenary Test

Here’s how I diagnose whether a company has a mercenary program or a real loyalty system.

Ask one question: Would your users leave the moment a competitor offered 10% better rewards?

If yes, you have mercenaries, not loyalists.

Companies with loyal customers almost universally operate on the opposite model. Apple doesn’t give discounts to keep customers; they charge premiums. Tesla fanatics defend the brand online without being paid to. Certain churches inspire lifelong commitment through shared belief systems. None of these organizations compete on transaction economics.

What they’ve done instead is establish Core Drive 1: Epic Meaning and Calling: a shared belief system, a sense of identity and belonging that exists outside any transactional calculation. Their customers aren’t comparing value propositions. They’re expressing identity.

No points program can compete with identity. It’s not even the same category of motivation.

5 Ways to Escape the Mercenary Trap

1. The CD1 Reframe: Make Rewards About Meaning, Not Money

When your reward economics can’t support generous payouts (which is most businesses, most of the time), reframe the value through Epic Meaning and Calling.

I worked with FullDive, a mobile browser company with a points-for-browsing system. The economics were brutal: users were earning pennies per hour of browsing. The math was humiliating when you ran it.

But then we discovered something: it costs only 5 cents to feed a starving child through certain aid organizations.

The repositioning was immediate. Instead of “Your browsing earns you reward points,” the narrative became: “Your browsing feeds starving children. Your points go to charity by default. You can scroll down to redirect them to yourself, but most people don’t.”

The monetary value didn’t change. Not a single cent. But the emotional experience transformed completely. Instead of “I worked all day and earned $2,” users felt: “I fed 40 starving children just by using my browser tonight.”

CD1 doesn’t require a good transaction. It requires a good story. And a good story is often available for free, if you’re willing to look for it.

2. Status Architecture: Make Levels Mean Something Real

I prefer status and leveling systems over exchangeable points, but only when the levels unlock real capabilities.

Every client gets this question from me: “What does it mean to be a top-tier member versus a new member? Do they have more powers? More access? More privileges?”

In well-designed games, leveling up makes you more powerful, more resourceful, more capable. In most corporate loyalty programs, reaching Gold tier means you have a gold-colored card. That’s it. There’s no functional difference in your daily experience.

That’s CD2 without substance. You’re creating the feeling of earning something without giving anyone anything real. Users figure this out quickly.

Octalysis Prime models this correctly: moving from one status level to the next unlocks real new features, interactions, and content that simply don’t exist at the lower tier. The level means something in the actual product experience.

3. Right-Brain Engagement Over Left-Brain Transactions

The easiest escape from the mercenary trap is to stop competing on the left-brain axis entirely.

CD5: Social Influence and Relatedness is extraordinarily powerful and almost free to implement. Facebook’s Like button didn’t cost users anything and required minimal technical investment, but it transformed the emotional experience of sharing content. I have a friend who won’t upload a photo unless he expects at least 5 likes. One-click appreciation is the real currency.

CD7: Unpredictability and Curiosity operates completely differently from predictable points. When users know exactly what they’ll earn, the math is easy and often unflattering. When there’s real surprise and delight, users can’t calculate whether the deal is bad. They’re just experiencing.

The highest-performing engagement systems I’ve studied share one characteristic: users pay the company, rather than the company paying the users. That’s the sign you’ve moved from Black Hat extraction to White Hat engagement. Users pay you to enjoy the experience. No transaction math required.

4. Cross-Product Value Exchange

When a company has multiple products or partnerships, the economics of reward programs can actually work, because you can give away something that costs you little but feels high-value to the user.

Microsoft does this reasonably well across Xbox, Edge, Bing, and Game Pass. A browsing reward that gives users access to Game Pass content costs Microsoft relatively little (marginal digital product costs approach zero) while feeling like a substantial benefit.

The Porsche Boneo program operates on the same principle within the Volkswagen Group family. Driving behavior earns points toward upgrades and experiences across a portfolio of brands. The abundance-to-company score is high; the perceived value to the user is also high. That’s the economic logic that makes reward programs sustainable.

5. Understand the Reward Spectrum: Earned Lunch vs. Free Lunch

In the Octalysis Framework, I distinguish two fundamental reward types that serve completely different psychological functions.

Earned Lunch (Game Technique #7): The reward is known, the requirements are known. Do X, get Y. Transparent, reliable, the backbone of most loyalty programs. Engages CD4 (Ownership) and CD2 (Accomplishment). Risk: if the deal is bad, users calculate it immediately.

Free Lunch (Game Technique #24): The user receives something valuable, normally requiring effort, without expected effort. Through luck, timing, or system design. Creates surprise and delight via CD7. Anchored against what others must labor for, making it feel special.

Free Lunch increases retention because it creates gratitude without obligation. Nobody calculates whether a surprise is worth it. You can’t over-justify a gift.

The best reward systems use both: Earned Lunch as the predictable progression backbone, unexpected Free Lunches as the delight layer that keeps users emotionally engaged between milestones. One makes users feel they’re progressing. The other makes them feel appreciated.

Frequently Asked Questions

Why do loyalty programs fail to create loyal customers?

Most loyalty programs fail because they trigger the Over-Justification Effect: small extrinsic rewards reframe an enjoyable customer experience as a bad transaction. When users earn points worth a fraction of a cent per interaction, their brains switch from enjoying the brand to feeling underpaid for their engagement. This reframe persists even after the program ends. The result is customers who stay only while the deal holds, then defect the moment a competitor offers marginally better terms.

What is the Over-Justification Effect in gamification?

The Over-Justification Effect occurs when introducing extrinsic rewards into an intrinsically motivated activity causes the brain to switch entirely to a transactional frame. The external reward “over-justifies” the behavior, making it feel like paid work rather than a choice. In gamification, this appears when points programs make users feel they’re earning pennies for their engagement. The Octalysis Framework addresses this through CD1 (Epic Meaning) and right-brain Core Drives that don’t create transaction economics at all.

What is the Mercenary Test for loyalty programs?

The Mercenary Test is a single diagnostic question: “Would your users leave the moment a competitor offered 10% better rewards?” If yes, your program has built mercenaries — customers who see the relationship as purely transactional. Brands with real loyalty don’t compete on transaction economics. They establish Core Drive 1: Epic Meaning and Calling, a shared belief system that makes price comparison irrelevant.

How do you build genuine customer loyalty without a points program?

The most durable loyalty comes from right-brain engagement: social recognition (CD5), surprise and delight (CD7), meaningful creative choices (CD3), and shared meaning (CD1). These don’t create bad transaction math because they don’t create transactions at all. Practically: give users public recognition, create unexpected Free Lunch moments, make status levels unlock real different capabilities, and connect your product to a mission bigger than the product itself.

Can any loyalty program actually work?

Yes, but the economics must be real, the status must unlock real capabilities, or the value must be delivered through cross-product exchange where marginal costs are low. The Porsche Boneo program and Microsoft’s cross-product rewards are examples of sustainable loyalty economics. The key distinction: are users engaging because the deal is good (mercenary) or because they value the relationship (loyal)?

WOULD YOU LIKE YU-KAI CHOU TO WORK WITH YOUR ORGANIZATION?