Acceptance Is the Asset, and It Is Not SoftwareWide moat

It is easy to underrate acceptance because it looks like plumbing. Visa's own description of its foundation layer states it plainly: connections to approximately 12 billion cards, bank accounts and digital wallets, and more than 175 million merchant locations across more than 200 countries and territories¹.

Each of those locations represents a terminal certified against a specification, an acquiring bank that has accepted the operating rules, a dispute process with defined liability, and a settlement arrangement that works in the local currency under the local regulator. None of that is difficult in isolation. All of it took decades, and the gating factor was never engineering but the willingness of two hundred national banking systems to say yes — one at a time, in their own languages, under their own laws.

This is why entrants with genuinely excellent technology keep arriving at the same conclusion: it is faster, cheaper and more certain to connect to Visa than to replace it. The consequence runs through the whole competitive picture. Most of the companies a consumer thinks of as payments businesses are Visa's customers, and their growth is Visa's growth.

The qualification is that the barrier is falling, not in the count but in the cost of adding to it. A phone can now be a terminal and an acquirer can be an API. That does not let anyone assemble 175 million locations quickly, but it does mean a rail that only needs the merchants who matter in one country can now do that work in months rather than decades — which is roughly the story of every domestic scheme that has succeeded.