There has been a lot of noise about AI across every domain, but let us be honest: AI is not a magic wand and it is not an external layer you bolt on. It is the ability to simplify complex decisions using data and technology, complementing human thinking rather than replacing it. I look at it through a very simple lens. AI helps you remember and respond. It does not help you matter.

As customers, we do not get attached to brands because they remembered our last purchase. We stay because the overall experience feels worth coming back to. So AI’s real role sits in what I call the 3R layer: remember what the customer does, recommend what might make sense next, and respond at the right moment.

Where it is overpromised is when we expect it to build loyalty. Loyalty lives in a different layer entirely, one driven by value, consistency, and trust. If that layer is weak, AI just helps you execute faster. Not better. I do not see AI as the hero. It enhances what is already there. If the foundation is weak, enhancement does not fix it.

The gap is not in AI. It is in an organization’s readiness to absorb AI. And the unfortunate truth is that most brands are not even clear on what they want from it.

The issue starts with the foundation. Most vendor conversations assume a clean world: unified data, aligned teams, clear use cases. Real organizations are far from that. The stepping stone for any organization is what I think of as a 4C Readiness framework: Clean, Connected, Contextual, Committed. Data needs to be clean. Systems need to be connected. Data needs context, not just volume. And teams need to be committed to actually using the outputs.

The biggest cost is not financial. It is cultural. Once teams lose trust in AI-generated outputs, adoption dies. The real work is unglamorous and happens entirely at the backend: fixing data, aligning teams, defining problems clearly. AI is the easy part. Getting the organization ready for it is the real challenge.

Absolutely, and customers are perceptive enough to sense it immediately. There is a meaningful difference between being known and being processed.

Imagine walking into a hotel where everything feels perfectly tailored but slightly mechanical. You know you are being handled by a system. But when a small, imperfect, human moment happens, when someone remembers something casually and without a script, it feels real. That is the distinction that matters.

The concept I keep coming back to is what I call invisible intelligence. If the intelligence is visible, the experience weakens. AI should work in the background, enabling people to deliver better moments, not replacing those moments. Hospitality is not about accuracy. It is about how an experience made you feel. Feelings do not come from perfect targeting. They come from genuine, well-timed human interaction that happens to be supported by insight underneath.

Honestly, the number is almost beside the point, because almost every organization believes it has arrived at AI readiness. The reality is that most are still congratulating themselves at the starting line. Most are in what I call the Data Illusion Stage: they have a lot of data, so they assume they are ready.

Readiness is not about data volume. It is about decision speed, agility, and clarity. Brands need to assess their data layer across three parameters: access, accuracy, and actionability. Can teams access the data easily? Is it accurate and consistent? Can you act on it quickly?

If a simple customer question takes multiple teams and several days to answer, AI will not fix that. When we built large data systems, the real objective was always to reduce the distance between signal and action. Organizations succeed when they focus on connecting their data rather than collecting it.

Customer behavior has never been linear. Brands have just forced it into linear models. As customers, we do not respond to a single message. It is always a combination: timing, context, mood, previous experiences. All of it becomes part of the decision.

So attribution, as we have traditionally understood it, is an oversimplification. I prefer thinking in terms of decision influence systems rather than attribution models. It is not about what caused the action. It is about what combination enabled it.

With AI now shaping content, timing, and targeting simultaneously, the need to shift becomes even more urgent. We need to move from credit-seeking to understanding-seeking, from last touch to full journey context, and from a funnel mentality to a flywheel one. Today AI is not just executing campaigns. It is co-creating the path. You cannot measure a journey with a single checkpoint.

Unredeemed points are one of the most misunderstood signals in the industry. From a business lens, they look positive. From a customer lens, they often mean the member did not find it worth engaging. No brand wants to hear that.

Redemption is a function of three things: value, simplicity, and timing. Is the reward worth it? Is the process easy? Is the offer relevant right now? AI can help with timing and nudges. It can simplify discovery. But it cannot create perceived value. If the reward does not excite the member or feels complicated to access, no volume of reminders will change behavior.

AI does not solve the underlying problem. It intensifies whatever already exists. If the program is strong, AI scales engagement. If the program is weak, AI scales indifference.

AI brings one uncomfortable truth to the surface: the customer experiences your organization as one entity, but you operate as many. Earlier, silos could survive because systems were slower and less connected. AI removes that buffer.

The simplest lens here is one customer, one memory, one experience. If different teams hold different versions of the same customer, AI will expose it quickly through inconsistent messaging, broken journeys, or irrelevant offers.

So the shift required is not purely technological. It is cultural. Organizations need to move from ownership of tools to ownership of journeys, and from measuring department success to measuring customer success. In my experience, alignment across people unlocks far more value than any AI model.

Global players win on scale of data. Regional players can win on depth of understanding. As a guest, what matters is not how much you are known, but how well you are understood in that moment. I think of this as context advantage.

A global system may know your travel frequency. A regional player might know you are traveling for a family occasion, and that changes expectations entirely. AI performs best when data carries meaning, not just volume. So instead of chasing scale, regional brands should focus on context, culture, and occasion.

In hospitality, relevance always beats reach.

For me, loyalty and CRM are not tech constructs. They come down to a very simple human question: did this brand make my life easier or better in any meaningful way?

Everything else, AI, data, platforms, is just infrastructure. Because in the end, customers do not stay because you know them well. They stay because you used that understanding wisely.