Jean-Philippe Poisson maps the AI strategies reshaping consulting delivery models
How major consulting firms are competing to control revenue, intellectual property, workflows and verification in the AI era
Jean-Philippe Poisson published a comparative map of how major consulting firms are approaching AI. His central argument is that these firms are not simply accumulating technology partnerships: each is trying to strengthen, defend or reposition a particular layer of its consulting delivery model.
The map distinguishes several strategic bets. Accenture is presented as turning generative and agentic AI into a measurable revenue stream. McKinsey and BCG appear more focused on converting internal knowledge, intellectual property and methodologies into scalable delivery infrastructure. The Big Four are shown embedding AI into specific business functions and workflows, while Bain is positioned more explicitly as an enterprise deployment channel for OpenAI models.
The useful insight is that the competition is no longer only about which consulting firm uses AI most effectively. It is about which layer each firm can continue to control: revenue, proprietary knowledge, workflow, distribution, executive relationships or verification. As models become more interchangeable, these surrounding layers may determine where durable value remains.
This matters beyond consulting. In Supply Chain Transformation, companies often depend on consultancies to frame processes, configure planning platforms, redesign operating models and translate analysis into execution. When AI industrializes research, documentation and parts of delivery, the scarce capabilities shift toward problem framing, domain judgment, integration with legacy systems and the ability to validate outputs before they influence operations.
The post also raises a client-side question: after an AI-enabled transformation, which capabilities remain under the organization’s control? A firm may accelerate delivery through external models and agentic workflows, but the client still needs ownership of process logic, evaluation criteria, data access, model changes and operational knowledge. Otherwise, faster delivery can create a new form of dependency.
Dataleo angle
This is a relevant Radar insight because the same strategic issue appears in Supply Chain AI projects. The durable value is rarely the model alone. It lies in the controlled layer around it: the decision workflow, business rules, operational context, validation process and connection to APS, ERP and BI systems.
For supply chain leaders selecting a consulting partner, the key questions should be explicit: who owns the resulting decision logic, who maintains the agents and evaluations, how outputs are verified, and whether internal teams can operate the solution after the engagement. The preferred outcome is not permanent dependence on a consultancy, but a governed Decision Architecture that the company can understand, challenge and maintain.