Gad Allon

The supply chain question used to be the pandemic. Now it is tariffs, and Allon's answer is a discipline rather than a forecast. Three worlds are possible: trade fully regionalized, a return to roughly what it was, and tariffs lower than today's. He asks which one an organization is built for, what the other two would cost, and tests the answer against the plants his own company runs. His verdict on reshoring: America has excellent engineering and poor manufacturing, because those plants' customers never asked them to be good at it. Audiences leave with a scenario plan they can build this quarter, and a clear read on what AI can and cannot fix on a factory floor.

Allon will not tell an organization which AI vendor to buy from. Before he presents to a chief executive, he spends the drive having two AI models play that executive back at him. Then he asks a room the harder version: if you are not doing that before you bring in your consultants, what are you even doing? He asks where an organization's own senior people come from once software does the first two years of the job, because third-year accountants do not grow on trees. He asks what is left of a website that a customer's software visits and the customer does not. Audiences leave with both questions put to their own plan, and with the standard he sets for anyone who wants to lead this rather than delegate it.

Your brand is not a moat, and Allon says so to rooms that would rather hear otherwise. Most organizations who believe they have a competitive advantage are describing something a competitor can copy. He uses the smile curve, from Acer founder Stan Shih, to show that defensibility sits at the customer end of a value chain or the research end, rarely the middle. His case is Nvidia, which had no moat for twenty years and then built one on purpose, next to the companies that did the same and were not lucky. Audiences leave able to say what actually protects their business, what only feels like it does, and where AI is dissolving the difference.

The instinct of his field is to add process. Allon removes it. Rules work for a quarter: everybody is clear, the organization speeds up, and then everything needs an SOP and it is slower than it started. His alternative is culture treated as decision-making infrastructure, not a values poster, because the point is that people make the right call when no senior person is in the room. His example is Pal's Sudden Service, a Tennessee drive-through chain that won the Baldrige Award with teenage employees who train a hundred hours before touching a burger. Audiences leave able to separate the rules that still earn their place from the ones holding the organization at its old size.

Getting bigger and scaling are not the same thing. Allon's SCALE framework looks at what will constrain the next stage, whether the organization is aligned and well led, and whether there is a predictable path to profitability. One piece is who is in the building. An organization needs bushwhackers who like the absence of a map, then drivers who can follow one, then people who run a faster lap than anybody alive. The three cannot stand each other, and his read is that most large technology companies have lost the first kind, which is why firms like Palantir and OpenAI outrun them. Audiences leave able to run the framework on their own organization and decide what has to change in how it operates, before growth forces the change.