Last week the Times of India published an essay called “Industrialised personalisation.” It traces a line from Frederick Winslow Taylor’s 1911 Principles of Scientific Management through Fred Harvey’s railway restaurants, White Castle, and the McDonald brothers rebuilding their kitchen around a nine-item menu. It arrives at a definition of personalization as a system set down so completely that anyone properly trained can replicate it and make the experience feel personal.
When the essay needs an example of that system working at the highest level, it turns to the Ritz-Carlton, and it gets there through my reporting. It cites me by name and draws on the research behind The New Gold Standard, one of the thirteen business books I have written about service organizations.
I am glad to be the evidence. However, I also want to be precise about what the evidence shows.
The essay is right about more than most executives will concede. Taylor’s logic did move out of the factory and into service, and it worked. The piece follows a new hire named Dan through his first shift at a McDonald’s, lost among the kiosks and the queue. Within a day you can tell he is new. Within a week he moves fluidly. Months later he is in a formal dining room and his thank you has stopped sounding rehearsed. His managers did their work.
Most service failures are system failures. The standard was never written down, or it was written and never taught, or it was taught and never resourced. Companies that skip the industrial work and go straight to talking about culture end up with staff who are warm and unreliable, and customers have a limited amount of forgiveness for that.
So build the system. Then understand where it stops.
The Ritz-Carlton example, from the inside
I spent years inside the Ritz-Carlton for The New Gold Standard. The guest preference database the essay describes is called Mystique, and it is as good as the essay suggests. A guest’s anniversary surfaces at the daily lineup. Champagne and a handwritten card reach the room before the guest does. Stay at one property and a waiter at another has already drawn on what the first one learned.
Mystique stores what a human being noticed. It has no sensing organ of its own. Someone overheard an anniversary mentioned in passing, understood that it mattered, and wrote it down. The system then distributes that noticing with a reliability no individual could manage alone. The noticing happens in a corridor, unprompted, by an employee who was paying attention and believed that paying attention was part of the job.
The company also gave every employee authority to spend up to $2,000 resolving a guest problem without asking permission. The figure gets quoted often. The design underneath it matters more. Leadership accepted that the person standing in the room would meet situations the manual had not anticipated, and decided in advance to trust that person’s judgment.
A system can be taught to execute. Judgment has to be permitted.
Why the boundary has become expensive
For most of the last century the boundary cost very little. Transactional service and attentive service arrived bundled in the same employee, so a company could buy one and receive some of the other without planning for it.
Artificial intelligence is unbundling them. Routing, order status, rebooking, a refund inside policy: that layer is being absorbed, and it should be. Nobody’s life improves by waiting on hold for a tracking number.
What stays on the human side of the line is the part the manual never covered. The noticing, and the judgment about what to do with what was noticed.
The ratio inside a frontline job has therefore inverted. That role used to be mostly execution with occasional judgment. It is becoming mostly judgment, staffed by people who are still hired, trained, and measured as though they were executing.
A test to run this week
Pull ten service recovery cases from the last quarter. Real cases with names attached, rather than a summary deck.
In each one, look for the sentence where an employee did something no policy required. Count how many of the ten contain that sentence.
For the ones that do, ask who taught that person it was allowed. A manager, a story from onboarding, a colleague they watched do it and survive. The answer is your actual authorization mechanism, and most organizations have never examined it directly.
If none of the ten contain that sentence, you have a system that executes well and a workforce waiting to be told what to do. That is a comfortable position until a customer arrives with a problem the system has not met before.
The question the essay leaves open
The Times of India piece closes with Dan leading you to the window table, where the sole arrives cooked the way it was in March. The system earned that ending.
The open question for anyone running a service business in 2027 is what Dan does when a guest says something the script has no line for. Whether he has been given permission to answer as himself. Whether anyone in the building will ever know that he did.
Write the standards down, teach them, resource them, measure them. Then spend what is left of your attention on the harder problem: the conditions under which a trained person decides to go past the training.
That decision is the product now.
Written by Joseph Michelli, Ph.D.
Have you read?
Essential Features Every Food and Beverage Maintenance Software Needs.
Empacer: Simplifying Customer Communication With Replies and Reminders.
Ranked: World’s Most Valuable Fashion Brands in 2026.
Dubai Property Investment 2026: Why Smart Investors Prioritize Selectivity.
Why America Needs AI Provenance Rules, Not Nationality Restrictions.
Follow CEOWORLD Magazine on: Google News, LinkedIn, X (Twitter), and Facebook.
Note: The views expressed are those of the authors and do not necessarily reflect those of CEOWORLD Magazine, its Editorial Board, or management. Content is provided "as is," may contain monetized links, and is not professional advice. See our Ethics & Guidelines for details. Reproduction requires prior written permission.
Contact info@ceoworld.biz for inquiries or media requests.
