The fantasy goes something like this. One day you will open your laptop, speak to an AI, and say, “Find me the perfect skiing holiday.” Seconds later, it will return a single answer. No friction. No thinking. Decision made.
It sounds elegant. It sounds efficient. And it is almost certainly wrong.
People do not decide like that. They never have. They never will.
Human decision-making is messy, comparative, emotional, and deeply contextual. We choose things not because they are objectively optimal, but because they feel right in contrast to other options. We need comparison. We need reassurance. We need to feel that we chose, not that a system chose for us.
This tension between efficiency and humanity sits at the very center of modern e-commerce. As AI, automation, and algorithmic optimization race ahead, the real risk is not technological failure. The real risk is succeeding at optimizing the wrong things.
Thinkers like Rory Sutherland, Richard Thaler, Roger L. Martin, and Nassim Nicholas Taleb have been warning about this for years. Their ideas are no longer theoretical. In a world where digital storefronts replace physical ones, their insights have become operational realities.
Why Nobody Clicks Google’s “I’m Feeling Lucky”
Google has offered the “I’m Feeling Lucky” button for years. Almost nobody uses it. That fact alone should make every e-commerce leader pause.
People do not want to be taken straight to one answer. They want to see options. They want to scan, compare, and reject before they commit. Choice itself is part of the experience. Remove it, and you remove confidence.
This is why AI systems that promise a single perfect recommendation misunderstand the problem. Humans cannot evaluate value in isolation. We only know what we want by seeing what we do not want.
E-commerce already reflects this truth, often without realizing why. Comparison tables convert better than single-product pages. “You may also like” sections increase basket size. Filters and sorting tools feel empowering even when they add friction. These features do not exist for efficiency. They exist because they mirror how the human brain works.
Contrast Is How Value Is Created
Real estate agents have known this for decades. Before showing you the house they want you to buy, they show you one that is slightly wrong. Often more expensive. Often less suitable.
When you finally see the intended property, it feels obvious. Better value. Better choice.
Nothing about the house changed. Your perception did.
This is not manipulation. It is contrast. And contrast is one of the most powerful forces in decision-making.
Online retail works the same way. A product displayed alone must justify itself absolutely. A product displayed alongside alternatives only needs to win relatively. That is an easier job.
The Efficiency Trap
Most businesses believe they are pursuing efficiency. What they are actually pursuing is measurable efficiency.
Cost reduction is visible. Value creation is not.
This is why cost-cutting initiatives look so good on spreadsheets and so bad in hindsight. Management consultants can always find costs to remove. The damage appears later, quietly, in lost trust, weaker brands, and lower pricing power.
Roger L. Martin captured this perfectly when he said that anyone can cut costs. What takes skill is cutting costs without destroying value.
The Dorman Fallacy, Digital Edition
Imagine a luxury hotel with a doorman. A consultant defines the role as “opening the door” and replaces the doorman with an automatic system. Tens of thousands saved per year.
Job done.
Except the doorman was also providing security, recognizing regular guests, helping with luggage, hailing taxis, signaling status, and setting the emotional tone of arrival. None of that appeared in the job description. All of it appeared in the guest’s experience.
Two years later, room rates fall. Loyalty declines. The hotel feels different.
In e-commerce, the doorman is customer support, post-purchase care, returns handling, and human judgment at moments of friction. When these are reduced to tickets per hour and average handling time, the spreadsheet improves while the brand quietly erodes.
Some Experiences Multiply
Not all customer interactions are equal. Some add value. Others multiply it.
A clunky checkout subtracts. A friendly, empowered human interaction multiplies.
This explains why a single great support experience can outweigh months of flawless logistics. You may interact with a brand digitally most of the time, but the one moment when something goes wrong defines how you remember them.
The Royal Mail story proves this. Areas with perfect delivery metrics were not necessarily loved. Areas with unreliable service sometimes were. The difference was the postman. A familiar face. A small favor. A brief human connection.
In e-commerce, where human contact is rare, its impact is amplified.
Trust Is the Ultimate Shortcut
Humans evolved to judge people long before they learned to judge systems. We have deep instincts for trust and almost none for evaluating operational efficiency.
When buying a second-hand car, most people cannot assess mechanical quality. So they assess the seller. A tidy home increases the price you are willing to pay. A seller in their underwear destroys it.
The car did not change. Trust did.
Online, customers use different signals, but the logic is identical. Tone of voice. Speed of response. Willingness to fix mistakes. These become proxies for quality.
This is why reviews, guarantees, and responsive support matter more than feature lists. They stand in for human judgment.
Feeling Good Matters
Richard Thaler’s idea of transaction utility explains why context changes value. A cold beer on a beach is worth more if it comes from a boutique hotel than a shack, even if you drink it in the same place.
The product is identical. The feeling is not.
In e-commerce, brands compete not just on products, but on how buying feels. Design, language, packaging, follow-up emails, and customer care all shape transaction utility.
Price wars happen when everything feels the same.
Marketing Is Bigger Than Ads
Advertising is a tool. Marketing is the system.
Marketing exists to find and keep customers profitably over time. Advertising helps, but it is only one lever.
Dyson’s success illustrates this. Its real advantage is not advertising. It is customer experience.
Inside Dyson, customer contact is treated as a privilege, not an inconvenience. Call centers exist to solve problems generously, not to clear queues quickly.
That mindset creates trust. Trust allows premium pricing. Premium pricing funds better service. The loop reinforces itself.
Automation Needs Boundaries
AI will transform e-commerce. That is not in doubt.
What is in doubt is how far automation should go.
The smartest systems do not remove humans. They reserve humans for moments where judgment, empathy, and flexibility matter most.
Self-service for the confident. Human support for the uncertain.
That balance is where trust is built.
Short-Term Numbers, Long-Term Damage
Public companies often struggle here because quarterly metrics dominate decision-making.
Founder-led and family-controlled businesses tend to think longer. They invest in employees. Employees treat customers better. Customers notice.
Brands like Costco, Dyson, and Enterprise demonstrate that looking after staff is not charity. It is strategy.
Differentiation Creates Value
Markets where everyone competes on the same dimension become hostile and unprofitable.
Differentiation benefits everyone. Companies earn margins. Customers gain real choice.
In e-commerce, differentiation rarely comes from features alone. It comes from philosophy, tone, and experience.
Marketing Is Fat-Tailed
Marketing is not linear. One idea can deliver value for years. Measuring everything by immediate ROI misunderstands this reality.
Treating marketing purely as a cost guarantees underinvestment.
The goal is not to be efficient everywhere. The goal is to be effective where it matters.
The Real Advantage
The future of e-commerce will not belong to those who automate the most, but to those who understand humans the best.
AI should remove friction, not meaning.
Technology should support judgment, not replace it.
In a world obsessed with optimization, remembering how people actually decide becomes a competitive advantage.
The spreadsheet is not the customer. The model is not the mind. And no algorithm can replace the feeling of choosing well.
The Part Everyone Forgets
The temptation, especially in e-commerce, is to believe that better data will eventually remove the need for judgment. That if the model becomes sophisticated enough, the human can step aside.
History suggests the opposite.
The more complex the system becomes, the more valuable human understanding becomes alongside it. Not instead of it. Alongside it.
People do not want to be optimized. They want to feel understood. They want to feel respected. They want to feel that when something goes wrong, there is a person somewhere who can see the situation as it actually is, not as a row in a database.
This is why brands built purely on efficiency struggle to create loyalty. Efficiency is invisible when it works and unforgivable when it fails. Humanity, on the other hand, is remembered.
The most successful e-commerce businesses in the next decade will not be the ones with the fastest automation or the most aggressive optimization. They will be the ones that understand where not to optimize. Where to slow down. Where to let judgment override process.
AI will get better at predicting behavior. It will not get better at caring why someone behaves the way they do.
That gap is not a weakness. It is the opportunity.
In a world obsessed with dashboards, funnels, and conversion rates, the real advantage will belong to the companies that remember a simple truth. Customers are not problems to be solved. They are people trying to make decisions.
And people do not choose what is perfect. They choose what feels right.
