Let me ask you something a little uncomfortable.
When a website greets you by name, recommends the exact running shoe you were eyeing, and remembers your shoe size, your last order, and the fact that you prefer Tuesday delivery — do you feel loved? Or do you feel handled?
Be honest. Because most of what passes for "personalization" in 2026 lands somewhere in that second bucket. It is competent. It is convenient. It is also, increasingly, identical to what the brand across the street is doing — because the brand across the street licensed the same recommendation engine from the same vendor you did. We have entered the age of what I call algorithmic intimacy: a thousand brands all whispering your first name at once, each one convinced that knowing your data is the same thing as knowing you.
It is not. And the gap between those two things is where the next decade of customer loyalty will be won or lost.
I have spent nearly four decades inside organizations that people are weirdly devoted to — Starbucks, The Ritz-Carlton, Mercedes-Benz, Airbnb, Zappos, and most recently One Medical. None of them earned that devotion by guessing your shoe size. They earned it by making you feel something. And as artificial intelligence drives the cost of personalization toward zero, that distinction stops being philosophical and starts being the whole ballgame.
Personalization Is Now Table Stakes — And Table Stakes Don't Win
Here is the trap. For fifteen years, marketers have been told that personalization is the loyalty strategy. Collect the data, tailor the offer, and devotion follows.
That was a defensible bet when personalization was hard. It is a losing bet now that it is easy. McKinsey's research on personalization found that roughly 71% of consumers now expect companies to deliver personalized interactions — and about 76% get frustrated when they don't. Read those two numbers together and the strategic reality jumps out: personalization has crossed from delighter to expectation. When 71% expect a thing, delivering it earns you no applause. It just keeps you in the game. The customer's reaction to a perfectly personalized experience is no longer "wow." It's "obviously."
I call this the personalization paradox: the better everyone gets at it, the less any single brand earns from it. AI is pouring gasoline on that fire. When every competitor can generate a tailored message, a custom bundle, and a name-checked email in milliseconds, "we know you" stops being a differentiator. It becomes the wallpaper of modern commerce — present everywhere, noticed nowhere.
So if personalization is now the price of admission rather than the prize, what buys loyalty?
Emotional Loyalty Is the Moat Algorithms Can't Cross
Here's the most important research a loyalty-minded leader can sit with this year. In Harvard Business Review, Alan Zorfas, Daniel Leemon, and Scott Magids reported a finding from years of customer emotion research that should reorganize your priorities: customers who are emotionally connected to a brand are roughly 52% more valuable than those who are merely "highly satisfied."
Sit with that spread. Not 52% more valuable than unhappy customers — 52% more valuable than your highly satisfied ones. The satisfied customer is the one your dashboards celebrate. The emotionally connected customer is the one who buys more, stays longer, forgives your mistakes, and tells other people about you. Satisfaction is rational; it lives in the head, and the head is endlessly persuadable by a competitor's better price. Emotional connection lives somewhere deeper and far harder to dislodge.
This is the distinction I want to plant in every loyalty conversation: loyalty is not retention. Retention is a customer who hasn't left yet — often because switching is annoying, or the contract isn't up, or they simply haven't gotten around to it. That's not devotion; that's inertia, and inertia evaporates the moment a frictionless alternative appears (and AI is manufacturing frictionless alternatives by the hour). Real loyalty is a customer who would switch easily and chooses not to, because something about you matters to them emotionally. One is a hostage. The other is an advocate. Algorithms are extraordinary at the first and nearly useless at the second.
The Starbucks Lesson: Devotion Lives in the Moment Data Can't See
When I was researching the experience that built Starbucks into a place people genuinely love, I kept expecting the answer to be the loyalty program — the stars, the tiers, the app. Those things matter. But they are not the source of the devotion. They are the scaffolding around it.
The source is smaller and stranger than any algorithm. It's the regular whose drink starts before they reach the counter. It's the barista who notices a customer is having a rough morning and writes something kind on the cup — a gesture no data field requested and no recommendation engine could ever generate. Those moments don't show up cleanly in a CRM. You cannot license them from a vendor. And they are exactly what turns a transaction into a relationship.
That's the uncomfortable truth about emotional loyalty: it is built in the precise places your technology is weakest. AI can tell you that a customer orders the same drink every Tuesday. It cannot, on its own, notice that this particular Tuesday they look like they could use a kind word. The data describes the what. The human supplies the meaning. And meaning is what people stay for.
What Earns Devotion When Everyone Has the Same AI
So how do you build the kind of loyalty an algorithm can't commoditize? Three disciplines separate the brands people love from the brands people merely tolerate.
Compete on feeling, not just fit. Most personalization optimizes for relevance — the right product, the right moment, the right offer. Necessary, but it's a rational play, and rational plays get matched. The brands that win loyalty layer an emotional objective on top: did this interaction make the customer feel known, respected, delighted, or cared about? Relevance gets you considered. Feeling gets you chosen again.
Use AI to free humans for the moments that carry weight. This is the move I'd put on the wall. Let your algorithms absorb the predictable, low-emotion work — the reorder, the status update, the routine recommendation — so that your people are present and unhurried for the high-emotion moments where loyalty is forged or lost. The complaint. The milestone. The first impression. The recovery after something went wrong. AI handling the rote work isn't the enemy of human connection; deployed well, it's what funds it.
Engineer for advocacy, not just repurchase. A retained customer buys again. A loyal customer brings you someone else. Those are different outcomes requiring different designs. Repurchase responds to convenience and price. Advocacy responds to emotion — people refer brands that made them feel something worth talking about. If your loyalty strategy ends at "they came back," you're measuring inertia and leaving the most valuable behavior — the referral — entirely to chance.
The Opportunity Hiding in Everyone Else's Sameness
Here's the part that should energize your next strategy session. The whole market is sprinting in one direction at once — toward more automation, more personalization, more algorithmic everything. Which means the market is converging. The experiences are starting to feel the same because, underneath, they increasingly are the same.
That convergence is leaving the most valuable ground in business wide open: the felt sense of being genuinely cared for by a brand that treats you as a person rather than a profile. As your competitors pour their budgets into out-personalizing each other on the rational layer, the emotional layer sits there, underdefended, available to any leader with the discipline to design for it deliberately.
So let me leave you with the question worth carrying into your next leadership meeting. If you stripped away your personalization engine tomorrow — the name-checking, the recommendations, the tailored offers — would your customers still feel a reason to stay? If the honest answer is no, then what you've built isn't loyalty. It's a very sophisticated form of convenience, and convenience is the easiest thing in the world for the next algorithm to beat.
Devotion was always personal. In the age of intelligent machines, it's also your most defensible advantage. If you'd like to pressure-test where real loyalty lives in your own customer experience, come find me at josephmichelli.com/contact.
