Executive Brief
Artificial Intelligence / Customer Proximity

When Your Customer Is a Machine

As AI agents increasingly search, evaluate, and transact on behalf of customers, businesses face a new strategic challenge: becoming visible, trusted, and relevant to machine customers. The conversation with Katja Forbes explores machine readability, distributed trust, brand signals, values, and the evolving customer journey as human intent and machine execution become increasingly intertwined.

Katja Forbes on how agentic commerce will reshape customers, brands, and competitive advantage

For decades, businesses have designed strategy around a seemingly obvious assumption: the customer is human. Technology might support the transaction, digital channels might change how products are discovered and purchased, and algorithms might influence what people see, but the ultimate decision has still been understood as a human one.

Katja Forbes believes that assumption is now becoming less reliable. As AI agents become capable of searching, comparing, filtering, negotiating and transacting on behalf of people and organisations, businesses are beginning to encounter a fundamentally different kind of customer. The human remains behind the transaction, but increasingly may not be the entity directly interacting with the business, evaluating the alternatives or executing the purchase.

Forbes sees the pace of this transition as particularly important for leaders. Developments that recently appeared to belong to a three-to-five-year horizon are moving towards a six-to-twelve-month window, creating a mismatch between the speed of the technology and the much slower rhythm of conventional strategic planning. Organizations therefore need to consider agentic commerce not simply as another digital channel, but as a change in the mechanisms through which customers discover, assess, and ultimately choose businesses.

 

From Machine Readability to Machine Preference

Forbes describes readiness for agentic commerce across three different altitudes: readability, trust, and values. Together, they provide a useful way of separating the technical requirements for participating in machine-mediated markets from the factors that may ultimately create differentiation within them.

 

The foundational requirement is machine readability. An AI agent needs structured product data, accessible information, APIs, schemas, and transaction capabilities that allow it to understand what an organization offers and interact with it without depending on the conventional interfaces designed for human customers. Much of the early work around agentic commerce is understandably concentrated at this level because a business that cannot be interpreted by machines risks being absent from the consideration process altogether.

Forbes does not, however, see machine readability as a sustainable source of advantage. As more organisations restructure information and transactions for machine consumption, these capabilities are likely to become basic infrastructure. The strategic challenge then moves to the second altitude, where an agent needs to establish whether an organization can be trusted and an organization needs corresponding confidence in the agent attempting to transact with it.

This introduces a different form of trust from the one businesses have traditionally tried to establish through customer experience and brand communication. Reviews, certifications, partnerships, and other forms of external validation become inputs that an agent can use when determining whether a supplier should be considered credible. Trust therefore becomes increasingly distributed across an ecosystem of signals rather than residing exclusively in the relationship between a company and an individual customer.

The third altitude is values, where Forbes sees the potential for more meaningful differentiation. An AI agent acting on behalf of a customer can potentially evaluate alternatives against environmental standards, ethical sourcing requirements, social criteria, and other expressed preferences. These factors can move from relatively broad statements of consumer intent into explicit constraints that a machine applies consistently when evaluating alternatives.

The implications are particularly interesting because human behaviour does not always follow stated preferences. A person may care about sustainability but still make a purchase based on convenience, price, or habit. An agent instructed to apply sustainability criteria can potentially enforce that preference systematically, making the clarity and credibility of an organization’s values more consequential to the purchasing process.

Forbes uses Patagonia as an example of how this might work. Its values are reinforced through signals beyond its own advertising and marketing, giving an agent a wider body of information from which to determine whether the organization aligns with a customer’s specified criteria. In this environment, organizations need to consider not only whether machines can understand their products, but whether machines can understand what the organization represents and find credible external evidence to support it.

 

Brand does not Disappear, but Machines Read it Differently

The prospect of machine customers creates an obvious challenge for traditional approaches to brand and marketing. Companies have invested heavily in customer journeys built around awareness, storytelling, aspiration, and emotional connection, yet an agent acting on behalf of a consumer may make a purchasing decision without meaningfully experiencing much of the brand environment created for humans.

Forbes’ argument is not that brand consequently becomes irrelevant, but that the sources through which brand exerts influence begin to change. An AI agent can evaluate product features, price, reviews, certifications, reputation, and alignment with a defined set of constraints without responding to emotional communication in the way a person does. What the organisation says about itself therefore exists alongside a much wider collection of signals that machines can use to determine what the organisation actually represents.

Forbes describes the resulting opportunity as an algorithmic moat. Rather than depending solely on proprietary technology, this form of advantage can emerge from the accumulation of external authority, trusted partnerships, credible reviews, certifications and consistent signals across the broader ecosystem. An organisation becomes easier for machines to interpret favourably because multiple independent sources reinforce the characteristics it claims for itself.

This creates a difficult adjustment for businesses accustomed to controlling their brand narrative. In a machine-mediated market, the company’s own description of itself may become only one input among many, while inconsistencies between corporate claims and external evidence become increasingly visible to systems capable of comparing large amounts of information.

The alternative is potentially greater commoditization. If an agent evaluates several organizations and finds little meaningful difference between them beyond basic functionality, price becomes an increasingly logical basis for optimization. Brand differentiation therefore remains strategically valuable, but maintaining that differentiation requires organizations to build signals that machines can identify and verify rather than relying exclusively on communication intended to persuade humans.

 

One Customer Journey Becomes Two

Forbes expects the transition towards machine customers to produce a hybrid commercial environment rather than the wholesale replacement of human customers. Some transactions will continue to be predominantly human, while others will involve agents at different stages of discovery, evaluation and purchasing. Organisations consequently need to understand how the familiar human customer journey connects with an emerging machine journey.

The human remains the source of the underlying need, preference or objective, while the agent increasingly interprets that intent and translates it into decisions and actions. Designing effectively for this environment requires businesses to preserve their understanding of the human context while also making the relevant information and transactions accessible to machines.

Forbes sees a potential role for what she describes as a human-machine bridge coordinator during this period of transition. Whether organisations ultimately formalise such a position is less important than the capability it represents: somebody needs to understand the human customer journey, identify where agents enter that journey, translate human requirements into machine-compatible interactions and ensure that the business remains accessible to both forms of customer.

The challenge may be particularly acute for smaller organisations. Larger companies generally have greater capacity to restructure data, develop APIs and redesign transaction systems. In contrast, smaller businesses can more easily assume that agentic commerce is an enterprise technology issue that can be addressed later. Forbes’ concern is that a lack of machine readiness could instead become a market visibility issue. If an agent cannot interpret a smaller supplier’s offering, establish its credibility, or interact with it efficiently, that supplier may never enter the consideration set regardless of the quality of its underlying product or service.

Agent readiness therefore reaches well beyond the technology function. Product information, reputation, customer experience, transaction architecture, and operating-model design all become connected because each contributes to whether machines can discover, understand, and confidently transact with the organization.

 

Leadership Needs to Get Closer to the Technology

The speed of agentic development creates a further challenge because many organizations are structured to respond to technological change through planning and delegation. A trend is identified, a business case is developed, responsibility is assigned, and a transformation program is established, often across a multiyear horizon. Forbes sees a growing tension between this approach and technologies whose commercial implications can change materially within months.

Her response is not that every executive needs deep technical expertise, but that leaders need enough direct experience with agentic systems to develop an informed view of their possibilities and limitations. Experimenting with AI tools, building simple agents, and observing how those agents search, interpret information, and make choices can give executives a much more tangible understanding of the strategic questions emerging around them.

Delegating that understanding entirely to technology specialists risks framing agentic commerce too narrowly as an IT issue. The more important questions concern how customers will find an organization, what information machines will use to evaluate it, how trust will be established, which forms of differentiation remain meaningful, and where value might migrate as machines assume more responsibility for purchasing decisions.

The infrastructure developing around agentic commerce reinforces Forbes’ view that these questions are becoming immediate. Payment networks such as Mastercard and Visa are enabling autonomous transactions, while ecosystems involving Google, OpenAI, Stripe and Shopify are developing capabilities around AI-mediated commerce. Forbes also points to developments in China, including experimentation within Alibaba’s ecosystem, as evidence that vertically integrated platforms may be able to move particularly quickly because discovery, evaluation and transaction can be connected within the same environment.

For strategic planning, the objective is therefore less about predicting the exact moment at which machine customers become dominant and more about identifying where machine mediation is likely to enter the customer journey. Organizations can then determine what would need to change if an agent, rather than a human, were responsible for finding the business, assessing its credibility, and deciding whether it belongs in the consideration set.

 

Competing When Machines Help Decide Who Gets Chosen

Forbes’ perspective on machine customers ultimately extends the AI discussion beyond the familiar focus on productivity and internal automation. Agentic systems do not only change how work is performed inside organizations; they have the potential to change the external mechanisms through which demand is interpreted and purchasing decisions are executed.

Businesses operating in that environment will still need to understand the humans whose needs ultimately drive commerce, but they will also need to become legible to the machines increasingly acting on those humans’ behalf. Machine readability provides entry into the market, trust determines whether an organization can be considered credible, and values and other forms of differentiation give an agent reasons to select one organization over another rather than defaulting to price.

For leaders, the practical challenge is to prepare for both customers simultaneously. Human intent and machine execution are becoming increasingly intertwined, which means the organizations best positioned for agentic commerce will be those capable of preserving genuine human relevance while expressing their products, credibility, and values in forms that machines can interpret. As Forbes’ argument makes clear, the strategic question is no longer confined to how businesses will use AI themselves. It increasingly includes how businesses will compete when AI is acting for the customer on the other side of the transaction.

 

To help you understand how AI and agentic commerce are reshaping customer behaviour, build stronger machine readiness, and stay ahead of emerging shifts, tune in to our podcast for practical insights and ideas to navigate the future of business.

 

Listen to “The Insurgent Mindset” podcast on Apple, Spotify, and Youtube.

Or contact us at mtice@insurgencegroup.com for more information.

Share this insight

Start a Conversation

Administrator preview. Public activation is pending GHL policy links and submission checks.

Loading the contact form…

Suggest a Topic or Guest