Brands no longer compete only for attention — they’re now competing for algorithmic interpretability too.
By Nuno Bernardo (DBA candidate, UBI Business School) and Gaston Fornes (Dean, UBI Business School)
The consumer was always human — until now
For decades, marketing was built on a relatively stable premise: brands competed to enter the mind of the consumer. From the original AIDA models to more recent versions of the consumer decision journey, the central assumption was that a person searched for information, compared alternatives, interpreted messages, formed preferences and made decisions. Even when the consumer journey stopped being understood as a linear funnel and came to be seen as a dynamic, digital and fragmented process, the unit of analysis remained the same: the human consumer.
That assumption is beginning to change. The emergence of AI agents capable of researching, comparing, recommending and even buying on behalf of individuals introduces a new strategic question: how do we do marketing when the decision-maker — or at least the decision filter — is no longer only a person, but an artificial intelligence system?
Share of model: the new metric brands are watching
A recent article in Harvard Business Review shows how LLMs and agents are transforming the way consumers research and buy. It uses the case of Pernod Ricard as an example: the company realised that some AI models were misrepresenting its brands and began monitoring what it calls share of model (SoM) — the frequency and favourability with which a brand appears in LLM responses, compared with its competitors (Acar & Schweidel, 2026).
This logic resembles Search Engine Optimisation (SEO), but goes further. It is no longer only about appearing in search engines, but about being correctly understood, classified and recommended by AI systems.
It is no longer only about appearing in search engines, but about being correctly understood, classified and recommended by AI systems.
In this context, it is possible to anticipate three types of interaction that are likely to coexist:
- Brand agents that interact with customers directly
- Consumer agents that act on behalf of individuals
- Full intermediation, where agents interact with other agents
(Acar & Schweidel, 2026)
The strategic questions this raises
This taxonomy forces companies to rethink their commercial strategy: Does my brand need its own agent? How do I ensure consumers trust it? And if consumers prefer to use their own agent, how do I make sure that agent chooses my brand?
Marketing for AI agents does not mean abandoning the consumer. It means accepting that the consumer–brand relationship will increasingly be mediated by autonomous systems. Brands that learn to remain relevant to humans while becoming legible to machines will be better prepared to compete in the next phase of e-commerce — and marketing more broadly.
A new form of literacy for executives
For professionals and executives, this implies developing a new form of literacy: knowing and understanding, for example, what is proprietary, what comes from third parties, what data is being used, who controls it, and — above all — how automated decisions are managed and audited.
For them, competitive advantage will not lie only in adopting AI, but in governing it; not only in automating decisions, but in understanding how those decisions are made; not only in speaking to people, but in designing brands that intelligent agents can understand.
Competing for algorithmic interpretability
The implications for marketing are profound. Brands — and the professionals behind them — will not only need to persuade consumers emotionally; they will also need to be interpretable, comparable and selectable by machines. This requires rethinking product information, ensuring semantic consistency across channels, making brand promises verifiable, designing integration points with external agents, and addressing many other challenges that are still emerging.
In this new environment, brands no longer compete only for attention, but also for algorithmic interpretability.
Looking ahead: the AI Journey
In conclusion, we know how consumers react to AI, but we still know very little about how AI systems discover, evaluate, select, purchase and recommend brands. The emerging concept is that of the AI Journey: understanding the consumer journey from the perspective of the agent, while keeping the consumer at the centre. Further research is needed to illuminate this phenomenon and its implications for marketing theory and practice.
References:
- Acar, O. A., & Schweidel, D. A. (2026). Preparing your brand for agentic AI. Harvard Business Review, March–April 2026. https://hbr.org/2026/03/preparing-your-brand-for-agentic-ai
Preparing Leaders for the AI Journey
These challenges are at the heart of how we’ve designed our artificial intelligence programmes at UBI, at both undergraduate and postgraduate levels. Our specialisations in AI & Machine Learning and AI in E-Commerce & Retail Marketing combine academic rigour, managerial orientation, and strong industry connections.
Highlights include:
- Specialisation tracks:
- AI & Machine Learning — developed in collaboration with Microsoft, with certifications and tools through Microsoft Learn for Educators
- AI in E-Commerce & Retail Marketing — developed in collaboration with Publicis Commerce, with access to Alexandr.ia, the AI-powered learning platform co-developed by Publicis and UBI Business School
- A HyFlex format designed for working professionals
- Small cohorts, applied projects, and strong networking opportunities
- The Venture Creation Lab
Training in AI is not simply about learning a technology — it is about preparing to lead organisations in a market where humans and intelligent agents will make decisions together.
Learn more about our AI programmes →
