Credence Good

Search goods can be compared before purchase, while experience goods reveal much of their quality through use, but credence goods remain hard to assess after use. A successful investment outcome does not prove an adviser chose the most suitable portfolio, and a loss does not prove the advice was bad.

The buyer may lack the expertise to diagnose the problem, evaluate a proposed solution or test whether the service was performed well, and the expert may be paid for the diagnosis, the treatment or both. These payment and information arrangements shape incentives, but they do not mean every expert behaves badly.

One risk is overtreatment, recommending a more complex or costly service than the client needs, while another is undertreatment, doing too little to address the actual issue. Overcharging is distinct again, because the bill may exceed the amount or scope justified by the work delivered.

The academic paper cited here models an expert who observes a customer's underlying loss and offers a repair, and in its stylised setting some useful repairs do not occur under the equilibria studied. This result illustrates how private information can distort choices, but it is not a measured failure rate for real financial advisers.

The paper also finds that more precise diagnosis need not always benefit the expert or lead to efficient decisions under its assumptions, so a manager should not convert that model into a rule to avoid diagnosis. The practical lesson is to ask what evidence supports the recommendation and what uncertainty remains.

The problem appears in many markets, since a business buying cybersecurity assurance, audit support or specialist valuation work faces a similar challenge: management may see the deliverable without being able to independently judge every expert choice. It should define a scope, request methods and compare recommendations before making a commitment.

For financial advice, ask for a written account of the goal, options considered, expected fees, material risks and assumptions, compare the proposal with a simpler alternative and ask how the adviser is paid. Credentials, regulatory status and references can help screen providers, though none guarantees a good recommendation.

Not every quality dimension is unobservable, since clients can often verify timeliness, courtesy, documentation and whether stated tasks were completed. The credence part is the harder question: whether the expert chose the right task, performed it competently and avoided unnecessary work.

A written scope and fee schedule help a buyer check what was delivered and compare options, and a second opinion can be worth the cost when errors are expensive.