Skeptics agree to pilot an idea they oppose, then starve it of resources or place it where it is least likely to work, and the engineered failure becomes evidence against the idea.
An example
Self-service buying for mid-sized customers was the proposal a product manager brought to a project-management software company's executive team: sign up and pay online, no sales rep. The head of sales, whose organization was built on rep-led deals, agreed to a pilot and picked the territory, large regulated enterprises in financial services, where every purchase went through procurement and security review. The pilot got no marketing budget, no product changes, and one part-time engineer. There was no comparison group. After six months it had converted forty-one accounts against a target of three hundred, and the head of sales presented the result as proof that self-service did not fit the company's customers. A board member asked why the test had not run in the mid-market segment the proposal was written for. Nobody had an answer, and the idea was shelved for a year anyway.
Why it works
The status quo, or the skeptic's own alternative, looks better without being sold, because the rival idea now carries fresh evidence of failure. Starving the pilot of budget, talent, or executive cover, or running it in an unrepresentative setting, makes the idea itself look worse. The failure loads political cost onto continued advocacy: the champion's judgment is questioned and future proposals are discounted. Cancellation becomes cheap, since the decision-maker can say the data made the call. It requires control over resources, scope, success criteria, and the test site, without that control being visible to neutral observers. Overt sabotage backfires, so the pilot must still look fair.
The thinking behind it
Lee Ross Ross's The Intuitive Psychologist and His Shortcomings: Distortions in the Attribution Process (1977) names the fundamental attribution error: observers explain outcomes by personal traits and underestimate the situation. A pilot that fails in the least suitable segment is blamed on the idea and its champion, and the conditions of the test drop out of the explanation. Naming the site choice publicly, before the failure story sets, puts the situation back into the judgment.
Jeffrey Pfeffer In Power in Organizations (1981), Pfeffer treats power as coming largely from control of resources, information, and the decision process. Whoever controls a pilot's budget, site, and success criteria controls the evidence the decision will rest on. That is why the defense treats the trial design as the real decision and asks for pilot governance by someone with no stake in the result.
Daniel Kahneman Tversky and Kahneman's Judgment under Uncertainty: Heuristics and Biases (1974) describes availability and anchoring: vivid examples dominate judgments of likelihood, and estimates adjust too little from a starting value. A single, concrete pilot failure becomes the anchor for what a full rollout would look like. Evidence from other settings and a comparison group supply a different starting point.
Warning signs
Pilot resourcing, scope, and site are set by the proposal's loudest internal skeptic. The test environment is the one least suited to the idea: the weakest tooling, the most adversarial management, the customers least likely to want it. No comparison group exists. Success criteria are vague or set after launch, and the result is quickly generalized to the whole initiative.
How to defend yourself
Treat the design of the trial as the real decision, and contest it before launch.
- Ask who chose the site, the budget, and the success criteria, and propose that they be set jointly or by a neutral party.
- Argue for a representative site, or several, chosen at random or by stratification, with a comparison group picked by someone other than the idea's opponents.
- Register the success criteria in writing before the pilot starts.
- Secure some resources of your own, so the pilot does not depend wholly on the skeptic.
If the pilot fails anyway, name the bias in how the site was picked, publicly, before the failure story sets, and bring evidence from other settings. An outside auditor reviewing the pilot conditions can reassign a failure to its real cause, and the organization's memory of a starved pilot helps the next proposal.
How a leader shuts it down
Assign pilot governance to someone with no stake in the outcome. Pre-register success criteria, choose sites by randomization or stratification, and have a neutral evaluator report results. When a pilot fails, review its conditions before concluding anything about the idea.