Learned-Skill Lock-InWide moat

Beneath the compatibility and the subscriptions runs the deepest and least visible thread of the productivity grip: learned-skill lock-in. A workforce fluent in Microsoft's products represents an enormous, invisible investment in training that walks in the door with every new hire and would have to be written off entirely to switch. Hundreds of millions of people know how to make Excel do what they need, and that knowledge is a cost of switching that never appears on any invoice.

The scale of the sunk investment is what makes it binding. Retraining a hundred thousand employees on unfamiliar software, and absorbing the lost productivity while they climb the learning curve, is a cost few executives will ever willingly bear — especially for tools that do roughly what the current ones already do. The alternative would have to be not merely better but transformationally better to justify the disruption, and it almost never is.

The lock-in renews itself with each generation of workers, because the fluency is taught in schools, in universities, and in every prior job, so that new hires arrive already knowing Microsoft's tools and expecting to use them. The training cost is borne by the wider world, not by Microsoft, and it deposits a pre-trained workforce on the company's doorstep for free.

For the owner, learned-skill lock-in is the cheapest moat imaginable: Microsoft did not pay to build it — the educational system and the labor market did — and it compounds automatically as more of the working world learns to speak the company's language. There is no cheaper way to own a market than to have taught it, in school and on the job, to think in your grammar. The trace it leaves in the accounts is selling cost: 8% of revenue last year against 12% five years earlier, on a business that nearly doubled in between¹.