Demonstrability, difficulty and persuasion: An experimental study of advice taking

Self-interested paid advisors should try to sell their solutions no matter how they came about.However, we present evidence that advisor persuasiveness depends on two dimensions of theirprior problem solving: solution difficulty and demonstrability. We report a laboratory experimentwith repeated advisor-client interactions where both these dimensions are independently varied.Persuasion rises in solution demonstrability and falls in difficulty. The reason is non-optimisingbehaviour: Advisors lacking in confidence fail to conceal difficult problem solving and thosereceiving their advice baulk when the proposed solution lacks objective success criteria irrespective of its promise. Our findings suggest differential prospects for persuasion and selling ofdifferent kinds of products, services and ideas