The law of parsimony, applied to business.

The law of parsimony, better known by its popular name Occam's razor, is a principle from philosophy of science and logic: when two explanations both account for the same facts, prefer the one that assumes less. It is named for William of Ockham, an English Franciscan friar and logician working in the early 14th century, whose writings applied the idea of preferring fewer assumptions to problems in logic and theology.

The line most often printed as "his" quote — that entities should not be multiplied beyond necessity — is not a verbatim sentence from Ockham's surviving texts. Later scholars condensed his position, expressed across several of his writings, into that one tidy Latin line. The sentiment is genuinely his. The specific wording is a paraphrase, and this page will not print it as a direct quote, because it is not one.

What the principle is not, is a claim that the simplest explanation is always the correct one. Reality is not obligated to be simple, and a genuinely simpler account that leaves out a fact the more complicated account explains is not preferred — it is just wrong. The razor is a tie-breaker, not a truth detector: between two explanations that both fit the evidence equally well, it tells you which one to prefer until new evidence says otherwise. Adding an assumption is a cost. It should buy you something.

Ockham was weighing explanations, not spreadsheets or software stacks. But the same test transfers cleanly to a company's systems, because a system is a kind of explanation: it is a company's standing account of how a fact gets from one place to another — how an order becomes a shipment, how a lead becomes a customer, how a receipt becomes a line on a closed set of books. Every extra tool, every extra hand-off, and every extra exception is an added assumption in that account, in exactly the sense the razor is skeptical of.

A company that runs the same process through four disconnected tools has not explained the process four times more thoroughly. It has added three more places the account can disagree with itself: three more logins, three more chances that the number in one system is not the number in another, three more people who have to remember which system is the real one this week. None of that complexity is buying the company anything the simpler version would not also buy it. That is the test the law of parsimony sets: not "how few tools can we use," but "does this additional part earn its place, or is it here because nobody removed it."

The three sections below apply that test to the parts of a company most operators recognize on sight: the systems that hold the data, the hand-offs between them, and the exceptions that grow around every rule if nobody prunes them. The examples are written generically, drawn from the shape of ecommerce operations rather than any one company's specifics — the order flow, the inventory count, and the support queue.