Undeclared Case Selection
Undeclared case selection is this publication's name for a structural defect in argument by example: a book's evidence is a set of cases, and the reader is never told how that set was assembled. Two filters usually do the work unseen. Cases are chosen because they succeeded, and cases are chosen from organisations the author is connected to. This is a property of the case set, not an accusation about anybody, and it is an editorial category rather than a research claim.
What it is
Each half of the problem has a name already. Selecting on the outcome is survivorship bias, which Phil Rosenzweig set out for business writing in 2007 under the heading of connecting the winning dots: study only winners and their common features read as causes. The second filter is undisclosed interest, which has a large literature in medicine and almost none in trade publishing.
What this term adds is the point about restatement. A connection may be visible somewhere in the book. The author may say in an acknowledgment, or in a chapter introduction, that a company is a client or that he sits on its board. What is missing is the declaration at the place where the claim is made, beside the growth figure or the turnaround story it qualifies. A disclosure forty pages away from the number it bears on is not doing the work a disclosure does.
The generic shape is easy to draw. A marketing book illustrates its method with six brand turnarounds. Three are former clients, one is a company whose board the author sits on, and none is a campaign that failed. Every one of those facts appears somewhere in the book, and none of them appears beside the figure.