We are often more certain than the evidence really allows.
More quotesComplex problems do not reward confidence. They charge it interest. Variation creates alternatives. Survivability stops one bad answer killing the search. Selection lets reality choose what management would otherwise keep for sentimental reasons.
A retailer pilots a layout in three shops. Product tests genuinely different onboarding routes with a small group. A manager tries meeting-free mornings for a month before declaring a cultural revolution. Each intervention is large enough to produce evidence and small enough to survive being wrong.
Remove one element and the method collapses. Variation without selection becomes an innovation zoo. Selection without variation crowns the first plausible thought. Experiments without survivability become reckless wagers whose owners must defend them forever because failure costs too much to admit.
Judge the portfolio, not each trial’s ability to become a case study. Define the loss boundary, feedback and closing date before launch. Make tests different enough to teach. Shut weak ones without converting their teams into cautionary tales. Failure should buy information rather than merely use budget.
The supposed safe alternative is often one enormous untested plan approved by everyone important. That avoids the embarrassment of several small failures and delivers one magnificent failure with branded lanyards. Uncertainty demands affordable mistakes, not ceremonial certainty.