When the Algorithm Becomes the Customer

• As algorithmic purchasing agents restructure markets, executives must understand the Shopper Schism, which separates humans who decide what they need from algorithms that execute purchases.
• Integrated executive education programs cover both traditional marketing strategies and algorithmic intermediation theory, which examines what occurs when AI controls customer access.
• The faculty who teach such programs must combine senior commercial experience with research credentials in emerging commerce theory.

 
Last year, a colleague working with a major global retailer shared something that stopped me cold. The company’s AI-powered procurement system had begun selecting suppliers based primarily on data quality and integration capabilities rather than brand strength or relationship history. The sales directors who had built those relationships over decades found themselves suddenly irrelevant to the purchasing decision.

“The algorithm doesn’t care about the 20-year partnership,” my colleague told me. Instead, it’s focused on structured data, the reliability of the application programming interface (API), and predictive demand accuracy.

This isn’t an isolated incident, and it isn’t science fiction. From Amazon’s anticipatory shipping patents to Instacart’s AI shopping assistants to the enterprise procurement platforms that autonomously manage supplier relationships, we’re witnessing a structural transformation in how commerce actually works. The entity making purchasing decisions is increasingly not human.

Yet walk into executive education programs at most business schools, and you’ll find curricula designed for a world where humans still control the buy button. We’ve become remarkably good at teaching executives how to use AI. We haven’t yet grappled with what happens when AI replaces the customer.

Understanding the Shopper Schism

For more than a century, the marketing theory taught in business schools has rested on a fusion of roles: The person who decides what is needed and the person who buys the product are the same human being. The entire architecture of the marketing funnel—awareness, consideration, preference, purchase—assumes this unity of consumer and shopper.

That fusion is breaking apart. I use the term Shopper Schism to describe the structural separation emerging across industries. Humans increasingly set preferences and constraints, which might include: I want coffee that’s ethically sourced, I need office supplies within this budget, I prefer hotels with these amenities. But algorithms execute the actual purchasing decisions.

This isn’t merely delegation. When consumers ask Alexa to reorder household essentials or procurement officers set parameters for AI purchasing systems, they’re not simply outsourcing tasks. They’re creating principal-agent relationships in which the agents—the algorithms—operate according to their own objective functions. Those functions might not align perfectly with human preferences, and they certainly don’t respond to traditional marketing tactics.

Could humans simply instruct the algorithm to weight relationships? In theory, yes. In practice, the algorithm operationalizes “relationship” as a data field such as supplier tenure or contract renewal rate, not as the tacit knowledge and mutual trust that give relationships their commercial value. More fundamentally, humans setting parameters often do not anticipate which criteria the algorithm will determine as decisive. The European retailer’s procurement team never imagined that API response times would outweigh 20 years of partnership. The algorithm did not go off-script. It revealed that the script was incomplete.

The implications are far-reaching, as I discuss in my forthcoming book, The Algorithmic Shopper. Brand equity, painstakingly built through emotional storytelling and cultural resonance, matters little to an algorithm optimizing for price-performance ratios. Sales relationships, developed through years of trust-building, carry no weight when purchasing decisions flow through automated systems. Even the marketing funnel itself—the conceptual backbone of customer journey mapping—loses coherence when there’s no human journey to map.

What’s Being Taught Versus What’s Happening

The widespread prevalence of AI is something that business schools are taking seriously. According to AACSB’s 2025 State of Accreditation Report, 97 percent of schools now reference AI use in their accreditation documentation. A 2025 report from AACSB examines how business schools are leveraging and teaching GenAI. And the Graduate Business Curriculum Roundtable found that roughly three-quarters of business schools currently teach generative AI in some form. 

But look closer at what’s actually being taught. The dominant framing treats AI as an efficiency tool that automates routine tasks, generates analytics for decision support, and enables chatbots to provide customer service. Students learn to deploy AI—but they don’t learn what happens when AI deploys itself as an autonomous market participant.

Today’s executive programs do not acknowledge that commercial capabilities are increasingly being directed at intermediaries who process information in fundamentally nonhuman ways.

Meanwhile, commercial excellence programs continue to operate in a parallel universe by providing rigorous training in pricing strategy, channel management, revenue growth management, trade marketing, and sales operations. These programs assume human buyers who respond to emotional appeals, negotiate based on relationships, and make decisions influenced by brand perception and sales skill.

The gap isn’t merely academic. I spent eight years as Group Chief Commercial Officer at De’Longhi, which manufactures coffee and espresso makers. In that role, I was responsible for commercial strategy across more than 120 markets, as well as executive development for senior commercial leaders across our global operations. This meant I evaluated and sponsored managers through programs at leading business schools worldwide.

Yet I never found a curriculum that addresses what we are seeing in our markets: algorithms increasingly making purchasing decisions that humans used to make. Today’s executive programs do not acknowledge that sophisticated commercial capabilities are increasingly being directed at intermediaries who process information in fundamentally nonhuman ways. The playbook that has worked for decades is becoming unreliable.

Alternate educational providers are coming up with their own playbooks. For years, several private executive education programs have been teaching versions of this material, often combining classroom instruction with structured immersions in Silicon Valley, Shenzhen, Shanghai, and other technology corridors where agentic commerce is being built. One example is the StartSe executive education platform in Brazil, where I recently delivered a keynote address on agentic commerce at the StartSe AI Festival 2026 in São Paulo.

These programs are not replacements for accredited graduate business education, but they are indicators of where senior executive demand is moving. They are nipping at the heels of traditional curricula precisely because they cover what is happening commercially before the academic literature has fully caught up. The challenge for accredited business schools is to close that gap before the gap closes the schools.

A Sample Integrated Curriculum

To address the disconnect between what’s happening and what’s being taught, business schools cannot simply add an AI module to existing commercial programs or sprinkle business cases into technology courses. They must design integrated curriculum architecture that treats algorithmic commerce not as a specialized topic but as the emerging context for all commercial activity.

In such integrated curricula, faculty could teach case studies centered on navigating the shift from human to algorithmic buyers; they could pair traditional commercial skills with algorithmic readiness assessments. Such programs could be relatively short. For instance, a three-day executive module might be structured as follows:

Day 1 establishes a foundation of commercial excellence, because executives can’t transcend what they haven’t learned. During this first day, they master pricing governance, revenue growth management, channel strategy, and trade investment optimization.

Business schools must design integrated curriculum architecture that treats algorithmic commerce as the emerging context for all commercial activity.

Day 2 introduces algorithmic intermediation theory, including the Shopper Schism concept and principal-agent dynamics. It also covers what I call Agent Intent Optimisation—an approach that structures commercial data so algorithms can find and evaluate it.

During this class, participants examine channel management through the lens of platform dynamics where AI intermediaries control customer access, and they revisit trade marketing fundamentals with the understanding that the “shopper” at the shelf may not be human. They gain further insights through case studies on B2B procurement that anchor abstract theory in operational reality.

Day 3 builds on a case study about platform intermediation to illustrate the concept of strategic integration. Instructors challenge participants to develop dual strategies, one for human decision-makers who form preferences and another for algorithmic systems that execute transactions.

Core elements of this module, including the Shopper Schism framework and algorithmic commerce theory, have been delivered to executive audiences at Harvard University’s Division of Continuing Education and at international executive forums, with strong reception. The integrated three-day program, along with a full 16-module syllabus, is now ready for deployment with academic partners.

Two Hypothetical Cases

Case studies could be among the most valuable learning tools in these modules. I have devised two hypothetical examples that demonstrate the power of the algorithm in commerce today:

Scenario 1: The Algorithm That Selects the Supplier

Global consumer goods manufacturer AlphaFoods has supplied a major European retailer for more than two decades. Its key account managers have built personal relationships with the buying office; annual negotiations are tough but predictable.

Then the retailer implements AI-driven procurement. Within months, AlphaFoods loses preferred supplier status on three product lines. The algorithm evaluates suppliers across dimensions the salesforce has never considered: data accuracy (product information completeness, API response times), supply chain reliability (on-time-in-full delivery rates), and price-performance ratios weighted against return rates. Brand heritage and buyer relationships aren’t scored.

For executives discussing this case study in a classroom, the learning opportunity is profound. They must grapple with what “selling” means when there’s no human to persuade. They also must answer complex questions: What capabilities would rebuild competitive advantage in this context? How could they optimize for algorithmic selection without abandoning brand investment?

Traditional management competencies remain necessary for the humans who set algorithmic parameters but are insufficient for the algorithms that execute decisions.

This scenario demonstrates that traditional key account management competencies—relationship-building, negotiation, category storytelling—remain necessary for the humans who set algorithmic parameters but are insufficient for the algorithms that execute decisions.

Scenario 2: The Platform That Captures Customer Loyalty

A mid-sized European hotel group has invested heavily in its loyalty program throughout the 2010s. By 2020, it has 800,000 enrolled members and is proud of its direct booking rate. Then, top executives examine the data more carefully. A high percentage of guests have made their reservations not through the membership program, but through Booking.com. Some choose the app because it’s convenient to use; some like the fact that it finds options with the best prices or amenities. What it doesn’t do is automatically refer customers to the hotel group’s properties.

This scenario lets executives explore what I call “loyalty transfer,” the shift of allegiance from brands to platforms when algorithms intermediate selection. The strategic question becomes: How can a firm compete for visibility when an algorithm is optimizing for platform revenue rather than brand loyalty?

An Uncommon Faculty Profile

The teaching of such integrated modules demands a particular type of faculty expertise. The instructor must understand both the operational realities of commercial execution and the theoretical frameworks emerging from research on algorithmic markets. Pure academics may lack commercial credibility; pure practitioners may lack theoretical grounding. What’s needed are faculty who’ve operated at senior commercial levels and who’ve engaged systematically with the scholarly literature on this transformation.

I’ll be direct: Such profiles are uncommon. The pathway from C-suite commercial roles to academic research positions isn’t well-trodden. Business schools that want to offer genuinely integrated programs might need to seek out nontraditional faculty candidates, such as practitioners who’ve chosen to complement their operational experience with rigorous research engagement. They also might need to rethink collaboration models, pairing research faculty with senior commercial practitioners in unconventional ways.

Business school administrators also might design faculty development initiatives that create structured pathways for practitioners with research interests—or for researchers seeking deeper commercial immersion. Finally, if they have relationships with executives who are interested in becoming professors, school administrators could encourage these individuals to attend one of AACSB’s Faculty Excellence Seminars, which are designed to improve teaching skills for instructors.

The Window for Leadership Is Closing

Major technology platforms aren’t waiting for executive education to catch up. OpenAI’s launch of Operator, Amazon’s continued evolution of anticipatory commerce, Google’s integration of AI shopping assistance into search—these aren’t beta experiments anymore. These applications are rolling out globally, reshaping commercial interactions at scale.

Executives trained exclusively in traditional commercial frameworks will find themselves increasingly disoriented. The skills that built their careers—relationship selling, brand storytelling, intuitive pricing—remain valuable but insufficient. Business leaders need new frameworks for a world where influence must be exerted on both human consumers (who still form preferences) and algorithmic shoppers (who execute purchases).

Business schools that move early to create integrated curricula will establish genuine thought leadership in a space that’s currently undefined. They can take this opportunity to do more than simply add courses; they can shape how an entire discipline evolves.

The question isn’t whether algorithmic commerce will transform executive education. It’s whether individual business schools will help define what that education looks like—or spend the next decade catching up to institutions that did.