Chipotle rebuilt their loyalty program this year. The marketing industry has been talking about it ever since.

You’ve probably seen the takes. “The blueprint for where restaurant marketing is headed.” “Every independent operator needs to pay attention.” “This isn’t a punch card anymore — this is the future of loyalty.”

I keep seeing these posts pushed into my LinkedIn feed. And every time, I have the same reaction.

The LAST thing any independent operator needs to do is copy anything a chain does.

Especially this.

What Chipotle built is not a blueprint. It is a confession. It is the perfected version of a game the chain was forced to play because it gave up the only game that actually mattered decades ago. And the marketing consultants selling it to independents as “the future” are running one of the oldest cons in the business: telling the person with the advantage that they need to abandon that advantage and adopt the chain’s workaround instead.

This post is going to take that position apart piece by piece. It is going to run long because the position is being sold everywhere and deserves an exhaustive answer. If you run an independent restaurant and you have been reading the Chipotle takes and wondering whether you need to catch up — this post is for you.

You do not need to catch up. You need to stop running in the direction the marketing industry is pointing.

Here is why.

What Chipotle Actually Built

Let’s start with the mechanism. Not the marketing language. The mechanism.

Chipotle’s new loyalty program does five things:

One. It scores every member on visit frequency and predicted lifetime value using an algorithm that watches their behavior over time.

Two. It uses that score to decide what offer to send each member, when to send it, and how much the operation is willing to spend to keep that member coming back.

Three. It delivers free food drops on a monthly cadence, but only to members who have “stayed engaged” — meaning the algorithm has determined the offer is worth the cost of the food based on the predicted return.

Four. It runs gamification and behavioral challenges — come in on a specific day, try a new item, hit a streak. Same mechanism as fitness apps and casino loyalty programs.

Five. It personalizes the delivery of all of this so each Guest feels like they are receiving something built specifically for them.

That is the system. Read it again without the marketing language.

Score members. Target retention offers. Time discounts to predicted churn moments. Gamify frequency. Personalize the packaging.

The chain is running an algorithmic frequency-extraction system with personalized discount delivery, dressed up in the vocabulary of loyalty. The word “loyalty” is being used to describe a system that has almost nothing to do with what loyalty actually is.

Which brings us to the first claim that needs to be dismantled.

“Points-Only Loyalty Is Basically Dead.”

The Paytronix report says this. The marketing consultants have picked it up. Points-only loyalty is dead. Chipotle killed it. The new thing is here.

Here is what actually happened.

Points-based loyalty was never alive.

It was always [Deal Loyalty] dressed up as relationship. A punch card told you nothing about the Guest and nothing about the operation. It rewarded frequency behavior that already existed and did nothing to change the underlying reason a Guest chose to return. Take the discount away and the Guest was gone the following week.

Everyone in the industry has known this for at least a decade. The Paytronix report is not news. It is a marketing document that gives the industry permission to sell the next generation of the same mechanism.

Chipotle did not kill points. They hid the points behind a smarter interface.

Members still earn points. Members still redeem points for rewards. The chain has added a layer of algorithmic decision-making on top that changes when and how members are prompted to earn and redeem — but the underlying mechanism is identical. Points in. Discounts out. Frequency behavior rewarded with financial incentive.

The Guest who chases the personalized offer is running the same play as the Guest who chased the birthday discount and the double-points weekend. The pattern is called [Deal Loyalty] and it has one defining characteristic: remove the deal and the Guest goes with it.

Calling this generation of Deal Loyalty “the death of points” is like calling a smartphone “the death of the telephone.” The mechanism is not dead. It has been re-packaged. Anyone selling you the “death of points” framing is either misreading the mechanism or hoping you will.

“A Free Food Drop Every Month If You Actually Stayed Engaged.”

The marketing consultants describe this as “an invitation to come back.”

It is not an invitation. It is a bribe timed to your predicted churn point.

Watch how the mechanism actually works.

The algorithm watches your visit frequency. It compares your current pattern to your historical pattern. It detects when your visits are slowing — when the interval between visits stretches, when your spend drops, when you skip a day you usually come in. At the moment the algorithm predicts you are about to stop being a repeat visitor, it triggers a “free food drop.”

That is not an invitation to come back. That is a retention intervention. The chain is buying back a visit they know they are about to lose from a Guest they never actually earned.

A real invitation is what a cast member does when they walk up to a Guest at Table 12 and say “we haven’t seen you and Karen in about a month — did we lose you?” A real invitation is a hand-written note in the mail that says “we miss you.” A real invitation is a phone call from the operator asking if everything was okay last time.

Those invitations do not require an algorithm because the operation is close enough to the Guest to know when the visit interval has stretched. The chain uses an algorithm because there is no one on their floor who knows any individual Guest by name.

The independent has that person. That person is called a cast member and their job is to know the Guests who sit in their section. Every operator I have worked with in forty-four years has had at least one cast member who did this naturally — a lead who recognized every regular by name, remembered their kids’ names, knew what they ordered last time and what they might want to try this time.

That cast member is worth more than any algorithm Chipotle will ever build.

And the operator who trades that cast member for an algorithm because the marketing consultant said it was “the future” has traded down. Traded the thing that cannot be copied for the thing every chain in America already has.

“Gamification And Challenges.”

Come in on a specific day. Try something new. Hit a streak.

Read that back to yourself.

The chain is designing behavioral challenges that train the Guest to modify their visit behavior to earn a reward. The mechanism is borrowed directly from casino loyalty programs and fitness app engagement systems. Both are behavioral engineering disciplines. Neither has anything to do with hospitality.

A Guest who comes in on Tuesday because Tuesday is bonus points day is not a loyal Guest. They are a Guest whose visit pattern is being shaped by an incentive structure that will disappear the moment the incentive changes. The relationship is not with the operation. The relationship is with the reward.

A Guest who tries a new item because a challenge told them to is not exploring the menu. They are running through a hoop.

A Guest who hits a streak is not building a habit around your operation. They are building a habit around the streak. The streak is the point. Your operation is the incidental setting.

Gamification is a Skinner box with a burrito at the end. The Guest performs the target behavior, receives the reward, and the operation congratulates itself for “engaging” them. What has actually happened is the operation has trained the Guest to jump through hoops for approval. The Guest is now less connected to the operation than they were before — because their motivation has shifted from wanting to be there to wanting the reward for being there.

That is not loyalty. That is behavioral conditioning.

And every independent operator who reads about gamification and thinks “we should do that too” is proposing to spend money teaching their Guests to be motivated by something other than the reason those Guests chose the operation in the first place.

Which is exactly backward.

“AI Watches How Often You Show Up And What Actually Brings You Back.”

This is where the language gets truly Orwellian.

The marketing consultants describe the AI system as if it were building a relationship with the Guest. “AI watches you.” “AI knows what brings you back.” “AI decides what to send you and when.”

Let me translate.

Surveillance targeting.

The chain is running behavioral surveillance on every member of its loyalty program. It is capturing visit data, purchase data, redemption data, engagement data. It is running that data through predictive models that estimate probability of churn, sensitivity to different offer types, likely response times, and predicted spend per visit. It is using those predictions to allocate marketing spend on a per-Guest basis.

None of this is a relationship. All of it is a targeting problem the chain is spending its way out of.

When a marketing consultant tells you “AI figures out what to send” — what they are describing is a scoring system, not a knowing system. The algorithm does not know the Guest. It knows a pattern of behavior and a probability distribution of responses. The Guest is a row in a database being optimized against a marketing budget.

Compare that to what an independent operator does when they know a Guest.

A cast member knows the Guest is celebrating their anniversary because they mentioned it three visits ago and the operation captured it in the [Guest History]. The lead knows the Guest is a wine drinker who leans toward Southern Rhônes because they have been served that Guest for two years. The kitchen manager knows the Guest is allergic to shellfish because it is flagged in the reservation system and reinforced with every visit. The operator walks the room and stops at that table because that Guest matters and the operator is on the floor to make that Guest feel seen.

That is knowing.

The chain does not know its Guests. It scores them. And the marketing consultants who confuse scoring with knowing are either not paying attention or hoping you will not.

The AI is not the future of loyalty. The AI is what the chain built because it gave up the ability to know its Guests when it scaled past the point where knowing was possible. The independent operator has not scaled past that point. The independent operator can still know their Guests. That is the entire advantage. And adopting the chain’s substitute for knowing is trading a diamond for the glass replica of a diamond.

“Everyone Gets Something Built For Them Instead Of The Same Blast As Everyone Else.”

Personalization is the newest word being used to disguise the oldest mechanism.

The claim is that Chipotle’s personalization moves it beyond the transactional. Each Guest gets a bespoke offer, timed and sized to their specific behavior. Therefore, the argument goes, it is not “just a discount” — it is a relationship.

This is nonsense.

A personalized discount is still a discount. Wrapping the same transactional mechanism in a bespoke package does not change the nature of the mechanism. It just changes the surface presentation.

The Guest is still being paid to return. The payment is still coming out of the operation’s margin. The Guest is still forming their reason-to-return around the payment. The moment the payment stops, the return behavior stops.

Personalization does not solve the [Deal Loyalty] problem. It hides the [Deal Loyalty] problem behind better production values.

And this is where the marketing consultants really benefit from the confusion. Personalized loyalty programs cost more to run than uniform ones. They require better data infrastructure, better modeling, better creative production, better delivery systems. The vendors selling loyalty stacks and demand systems and AI-driven retention platforms are selling the personalization layer at a premium. If independents believe personalization is meaningfully different from generic discounting, independents will pay more for it. Which is exactly the outcome the vendors are working toward.

Do not fall for it.

A discount targeted at you specifically is still a discount. Being told the discount was chosen just for you does not change the mechanism. It changes the marketing copy on the way to your inbox.

“Scoring Members On Visit Frequency And Predicted Lifetime Value.”

Now we get to the metric that ties the whole system together.

Predicted lifetime value is a Road 1 counterfeit of actual lifetime value.

Here is the difference.

Actual [LTV] is the compounding value realized through relational investment across the full arc of the operator-Guest relationship. It is built one visit at a time by a cast that knows the Guest, an operation that captures Guest preferences in a Guest History, a systematic follow-up cadence that deepens the relationship over time. Actual LTV emerges from architecture. It is produced by the operation. It is realized as revenue but it exists first as relationship.

Predicted LTV is a statistical model. It takes visit patterns, spend data, and behavioral signals and outputs a probabilistic forecast of how much revenue this Guest is likely to produce over the next N years. It is used to decide how much marketing spend to allocate to retaining them. Predicted LTV is consumed by the operation. It is realized as a targeting decision. It does not exist as relationship. It exists as a number.

These are not two different tools measuring the same thing.

They are two different mechanisms producing two different outcomes.

Actual LTV is what the operation builds when it invests in the architecture of the Guest relationship. Predicted LTV is what the operation calculates when it has to decide how much to spend to hold onto a Guest it never actually created.

The chain uses Predicted LTV because it has to. It cannot build actual LTV because it does not have the architecture that produces actual LTV. It has no cast that knows the Guest. It has no Guest History that captures who the Guest is across visits. It has no relational compounding mechanism because the operation is not close enough to any individual Guest to compound anything.

The chain’s Predicted LTV number is a proxy for actual LTV that the chain will never build.

Independent operators who adopt Predicted LTV thinking are adopting a proxy metric for a thing they can actually produce. They are trading the real asset for the statistical estimate of the asset. This is a downgrade dressed as sophistication.

And the marketing consultants selling Predicted LTV analytics to independents are selling them a chain instrument that the independent’s own architecture makes unnecessary.

If you know your Guests by name, capture their history, deepen your relationship with them visit by visit, and structure your cast to serve them relationally — you do not need a statistical estimate of their lifetime value. You already have their lifetime value. It is walking through your door on Thursday night.

“This Isn’t A Punch Card, It’s A True Loyalty Program That Targets, Develops, And Rewards Frequency.”

This is the sentence that gives the whole game away.

“Rewards frequency.”

That IS the punch card. That has always been the punch card. Adding AI does not move it off the punch card. It just makes the punch card personalized, predictive, and prettier.

Frequency is not loyalty.

Frequency is habit.

Loyalty is what happens when a Guest returns to your operation because they were changed by the experience of being there — because something about your operation matters to them in a way that transcends the transactional exchange. The Guest who is loyal is not counting punches. They are not chasing rewards. They are not optimizing their behavior to unlock the next tier of the program. They are returning because the operation has become part of their life, and the operation returns that recognition every time.

That is loyalty.

Habit is what happens when a Guest returns to your operation because it is convenient, familiar, and rewarded with incentives that make the return marginally cheaper than alternatives. Habit collapses the moment a better incentive appears elsewhere. Loyalty does not.

Chipotle is rewarding habit. They are calling it loyalty because “habit program” would not sell as well. The marketing consultants are amplifying the naming choice because “loyalty” is a word with cultural weight and “frequency habit” is a word with mechanical weight. The name is doing work the mechanism cannot do.

Do not confuse them.

The Guest who visits your operation four times a week because you are the closest sandwich place to their office is not loyal. They are convenient. Move a Jimmy John’s across the street and you will find out fast.

The Guest who visits your operation once a month for their family birthday dinner because your operation has been part of every family birthday dinner for the last eight years — that Guest is loyal. Move a Jimmy John’s across the street and they will keep coming to you. Move you across town and they will follow.

Frequency does not tell you the difference. Frequency is a single-dimension metric that cannot distinguish habit from loyalty. Chipotle is running a frequency-optimization system. That is fine for Chipotle. Chipotle does not have the depth of relationship that would allow them to distinguish anyway.

The independent operator does have that depth. Or should. Which means the independent operator is measuring the wrong thing when they measure frequency alone.

“Built Around What They Wanted Guests To Do Next And How To Make That Effortless.”

Read that sentence carefully.

“What THEY wanted GUESTS to do next.”

The operation decides what it wants. The Guest is the subject being acted upon.

That is not hospitality.

Hospitality is built around what the GUEST wants to experience next and how to make THAT effortless for the Guest. The direction is opposite. The operation’s job in hospitality is to serve what the Guest wants, not to engineer what the operation wants from the Guest.

Chipotle’s system inverts that entire relationship. The operation identifies the target behavior (increased visit frequency, higher predicted LTV, gamification challenge completion) and engineers the Guest experience to produce that target behavior. The Guest is not being hosted. The Guest is being conditioned.

This is behavioral engineering with a marketing budget. It is a completely coherent business strategy for an operation that has no other option because the operation has no cast, no floor presence, no relational architecture, and no way to compete on hospitality. Chipotle cannot deliver hospitality at scale. So Chipotle has stopped trying and switched to behavioral engineering.

Fine. That is their choice. It is the right choice for them given the constraints they operate under.

But telling an independent operator that behavioral engineering is “the future” is telling them to abandon the one thing they can do that Chipotle cannot.

The independent operator can deliver hospitality. Actual hospitality. Not scaled hospitality. Not personalized-at-arm’s-length hospitality. Not algorithmically-triggered hospitality. Actual hospitality — the human interaction between a cast member who is trained and present and a Guest who arrives and is recognized and served.

That is the independent’s advantage.

Trading it for behavioral engineering is not modernization. It is capitulation.

“Is Your Program Actually Changing Behavior? Or Is It Just Rewarding Stuff People Were Going To Do Anyway?”

This is the closing rhetorical move in every one of these posts. It sounds like a challenging question. It is meant to make the independent operator uncomfortable and feel like they need to upgrade their thinking.

The question is wrong.

The right question is not whether your loyalty program is changing Guest behavior. The right question is whether your operation is building the kind of relationship with your Guests that makes a loyalty program unnecessary.

Behavior change is what you engineer when you cannot produce the return through the experience itself. If the experience produces returning Guests naturally — because those Guests were changed by the experience, because those Guests feel a genuine connection to the operation, because those Guests want to return for reasons that have nothing to do with the incentive structure — then you do not need a program to change behavior. The Guest is returning without one.

The chain uses loyalty programs because the chain cannot produce that experience. The chain is running a scaled operation with commodity food, uniform service delivery, and no relational depth per Guest. Something has to close the gap between “reasons the Guest chose us the first time” and “reasons the Guest chooses us again.” For the chain, that gap is closed by the loyalty program. Points, personalization, gamification, targeting — all of it is architecture for closing a gap that exists because the chain has nothing else that would close it.

The independent operator has something else. The independent has the ability to make the experience itself the reason for return. Which means, if the independent is doing their job right, the gap does not need to be closed by a program.

The Guest who returns because they were changed does not need to be scored, targeted, gamified, or bribed. They just need to be known. By name. By the cast. By the operation.

That is not a program. That is an operating discipline.

Ask a different question.

Is your operation producing returning Guests because the experience is worth returning to? Or are you propping up the frequency numbers with discount mechanics that hide the fact that the experience is not doing the work?

That is the actual question. And it does not have a personalized-AI answer.

The Consultants’ Frame

Before we close, I want to say something about the people selling this position.

The Chipotle takes are not neutral industry observation. They are being written by people who sell demand systems, loyalty stacks, AI-driven retention platforms, and behavioral engineering services to independent operators. Their business model depends on independents believing that chain-style algorithmic loyalty is the future — because that is the product they sell.

When a marketing consultant tells you Chipotle is “the blueprint for where restaurant marketing is headed” — read that sentence carefully.

They are not describing where restaurant marketing is going. They are describing where their business is going. Those are not the same sentence.

The vendor selling loyalty tech benefits when independents adopt loyalty tech. The vendor selling demand systems benefits when independents believe they need demand systems. The vendor selling AI-driven personalization benefits when independents believe personalization is the future. There is no consultant selling relational architecture consulting as an alternative because relational architecture is not a product you can package into a monthly SaaS subscription. It is an operating discipline that the operator has to build themselves, in their own operation, with their own cast, over years.

That is inconvenient for the marketing industry. So the marketing industry does not talk about it.

You should.

Every time you see a Chipotle take pushed into your feed, ask what the person writing it sells. If the answer is loyalty stacks or demand systems or retention platforms — you are reading marketing copy for their business, not analysis of yours.

That does not make them wrong about everything. It does mean their frame is not neutral. And their frame is being sold to you as if it were.

The Independent’s Position

Here is what you should actually do.

Do not copy anything Chipotle just built. Not the algorithm. Not the personalization layer. Not the predicted LTV scoring. Not the gamification. Not the AI-triggered offer delivery. None of it.

Instead, build the thing Chipotle cannot build.

Know your Guests by name. Train your cast to recognize repeat Guests within three visits. Build a Guest History system that captures who each Guest is across visits — occasions, preferences, milestones, allergies, wine leanings, everything the cast needs to make the next visit better than the last one. Run a follow-up cadence that stays in touch between visits without asking anything in return. Walk the floor. Stop at tables. Ask questions. Remember answers. Feed everything back into the system.

That is your operation.

That is [Relational Compounding] running as designed. Visit by visit, the relationship deepens. The Guest returns not because they are being scored or targeted or bribed, but because they have been changed by the operation and want the operation to remain part of their life.

That is loyalty. Actual loyalty. Not the redefined version the marketing consultants are selling.

And it is the one thing the chain physically cannot copy back. Chipotle cannot install a cast in 3,600 locations that knows every Guest by name. Chipotle cannot build the depth of relational architecture that an independent can build in a single restaurant. Chipotle is not competing with you on this axis. Chipotle abandoned this axis when it decided to scale.

You still have the axis. Use it.

That is the play. It has always been the play. Every marketing trend that tells you the play has changed is a marketing trend selling you something else.

What Changes Tomorrow

Look at your operation and ask three questions.

Do we know our Guests? Not “do we have their email addresses in a database.” Do we — the operator, the cast, the room — know them. Would a cast member recognize any of our regulars by name, know what they ordered last time, know why they come?

Are we capturing what we know? Every restaurant knows something about its Guests. The question is whether that knowing is captured in a system that survives cast turnover, shift changes, and the operator’s memory. If the knowledge lives only in the head of one server who might leave next month, the operation does not actually know its Guests.

Are we deepening the relationship visit by visit? Or are we treating every visit as a transaction — the food, the check, the goodbye, the next Guest? If every visit is a fresh transaction, no compounding is happening.

If the answer to any of these questions is no, that is your work. Not the loyalty program. Not the AI. Not the personalization layer. The actual architecture that makes the loyalty program unnecessary.

Start there. Ignore the Chipotle takes. Ignore the consultants selling you shortcuts. Build the thing only you can build.

The chain’s future needs algorithmic loyalty because the chain has no other option. Your future does not need it because you do have other options.

Choose accordingly.