Narrow positioning gets blamed for stalled expansion, but the narrowness is rarely the problem. A beachhead works precisely because it is small enough to dominate. What determines whether you ever leave it is something founders decide almost carelessly at the start: whether the segment they picked has neighbors. A niche has neighbors when the infrastructure, vocabulary, and buying process you build for it transfer to the segment beside it. A niche has none when everything you learn is specific to the people you learned it from.

Amazon sold only books for its first three years. Books were not chosen for margin, since bookselling is a famously thin-margin business. They were chosen because the catalog was enormous, which forced the company to build search, recommendations, warehouse logistics, and a customer review system that no physical store could match. Every one of those assets was indifferent to what was inside the box. When music and video arrived in 1998, the company did not rebuild itself; it pointed existing machinery at a new list of items. The books were the beachhead. The logistics were the landmass.

The trap looks identical from the inside whether or not the neighbors exist. A team serving veterinary clinics can build deep workflow knowledge, a specialized vocabulary, and a sales motion tuned to one buyer, then discover at expansion time that none of it means anything to a dental practice, because the regulatory frame, the billing path, and the purchasing committee are all different. Those founders did not niche too hard. They niched onto an island. The question worth answering before committing is not how big the first segment is, but what the second segment would need to have in common with it for the first one's assets to carry over.