Main idea

In a wiki of reading notes on marketing, you ask a graph tool for the shortest path between two notes that have nothing in common: Cialdini's rule of reciprocity, a principle of social psychology, and churn in a subscription business. It finds one in eight steps: initial commitment that produces consistency, high risk that builds trust, a high price as a signal of quality, switching costs that reduce price sensitivity, retention as a growth multiplier, and finally churn as a death spiral.

Nobody wrote this line of reasoning. Each link was laid down separately, note after note, book after book, with its own justification. Strung together by the computation, those links form an argument that runs from the psychology of commitment to the economics of retention.

The path therefore invents nothing: it assembles relations already written, in an order the author had never walked through.

Why it matters

This makes the bringing together of distant ideas operational, when it is usually a matter of chance or memory. You can ask for it, between two chosen notes, and get an argued lead rather than an analogy.

It also makes the quality of the links a condition for the quality of the reasoning obtained: a path made of explained links reads like an argument, a path made of silent links does not read at all.

Nuances and limits

The shortest path is not the best argument. It readily goes through the most connected notes: the path between the chasm and willingness to pay goes through three reading maps, which is accurate but teaches nothing.

And a plausible path still has to be checked: each link in the chain is justified locally, nothing guarantees that the whole chain holds.

Open questions

• How do you tell a path that reveals a real relation between two ideas from one that merely chains neighbourhoods together?