How To Pick A Quack: Theory

A few weeks ago I posted on the 25 most common types of clues mentioned in “How to pick an X” web guides for these 18 types of experts (sorted by ave # clue types per guide):

Lawyer, Private Eye, Therapist, Accountant, Chiropractor, Auto Repair, Doctor, Music Teacher, College, Dentist, Financial Advisor, Interior Decorator, Astrologer, Hedge Fund, Pastor, Charity, Health Plan, Fortune Teller

I said:

Guides do not often mention outcome-related clues, presumably as few customers attend to them. In general, we can’t tell if a type of expert X is a “quack”, where “better” versions don’t help customers much more with outcomes, by the kind of clues people use to pick X. Maybe most people can’t tell the difference.

Let me elaborate here, and describe what sort of expert incentives are produced by customers using each type of clue. (Next to each clue is the % of guides mentioning that clue.) A key question to keep in mind is how, if at all, use of that clue plays out differently if this type of expert is a “quack” who provides no more customer value than reassuring customers that they cared for, are high status, and are doing what most people think is what you are supposed to do for their problem.

Specialization (66%) – “Differentiate your product” is a powerful marketing principle. It gives you more customer loyalty and market power. If customers believe your type of product gives value, and that it varies, your messages picking them out as especially well-matched to your product will seem especially credible to them. After all, you give up on all those other customers just to match with them. This all gives experts strong incentives to vary their products, regardless of other issues.

Price (61%) – No matter what they are getting, customers know that all else equal that they want to pay less. This gives experts an incentive to cut prices when they can. Except that customers often infer that higher price products deliver more value, even when they don’t.

Credential, License (49%) – Customers are reassured to see that experts are supported by governments, universities, and other large long established organizations. Sure one con-artist might try to fool them, but only a “conspiracy theorist” would believe that all these big organizations would go along with a con that continues for decades. And even if they are conned, that doesn’t look so bad if most everyone else is conned with them. After all, they say no one ever got fired for buying IBM. So experts seek to organize to endorse each other, and to gain government endorsement. Usually that just costs money.

Customer Referrals (47%) – Customers who are worried if they can trust experts are especially reassured by endorsements from people in their social world. The problems of these local people are more similar to their own problems, and such local folks give them local allies who makes it less likely that they will be accused of being a fool by other local folks. I can’t be making such a bad mistake if many others in my social world are doing the same thing. This gives experts incentives to locate themselves visibly within a connected group of customers, and to try to attract large enough fractions of each social world.

Gut Comfort (44%) – People who feel intuitively comfortable with an expert are less likely to accept skeptical doubts about them. And if you feel comfortable, other people probably do as well, making it less likely that they will accept skeptical doubts. Many people are willing to ignore much concrete evidence that conflicts with their direct intuitions. Experts thus have incentives to act in ways that people like and feel comfortable with. Experts are pushed to pander to customer preconceptions, and also to be pretty, charismatic, etc.

Experience (40%) – If experts actually knew something, they’d probably learn more with more experience. So it makes sense for customers to prefer experts with more experience. The more prior customers an expert has helped, the more social proof there is that others have accepted their expertise. So experts want to collect and advertise their experience.

Online Listings (34%) – All else equal, costumers would like to quickly find and choose an expert. And any expert included in some official list online can’t be a complete con-artist, right? After all, if they were wanted by the police, the police could find them via an online listing. So for customers who feel in a rush, quickly browsing an online list and grabbing one can be awful tempting. So for experts, adding their name to an online list seems an easy win.

Disciplinary Actions, Complaints, Crimes (31%) – These actions tend to be about pretty extreme malfeasance, such as stealing, lying, sex, disappearing, etc. To avoid such outcomes, a customer is naturally wary of picking an expert for which there are official complaints or punishments. Not just to avoid such things happening to them, but also for fear that if anything else goes wrong, they’ll be blamed for not heeding these warning signs. This will push experts to avoid getting caught for such obvious stealing, lying, etc., though that usually has little to do with whether they actually give much value to customers in ordinary cases.

Location (27%) – All else equal, it makes sense to pick an expert located closer to you. Not only will this make it easier to travel to see them, they are more likely to be integrated into your social world, offering social proof to others in that world that you haven’t made a terrible mistake by choosing them. So experts seek to locate themselves near customers.

Expert Referrals (25%) – Customers are assured to see an expert endorsed by other customers, the government, and by associations of the same type of expert. But in addition, customers are also assured to see experts endorsed by experts of other types. This induces experts of different types to coordinate to endorse each other.

Professional Associations, Better Business Bureau (24%) – The more people and groups who endorse an expert, the more social proof a person has backing that expert, and the harder it will be to accuse them of making a bad choice. So experts of each type try to coordinate to create associates that will endorse them.

School (24%) – We feel assured to pick experts who have the marks of prestige, and in our society schools are one of our most widely recognized such marks. Human foragers spent a million years learning to copy and follow the prestigious, and we know it will be harder for others to blame us for picking more prestigious experts. Experts thus seek prestige markers, like school degrees.

Track Record (18%) – This is the first item on this list that has anything to do with whether customers actually get value from hiring experts, aside from in extreme failure scenarios. Alas, official track records are often edited to give misleading impressions. For example, the cases with the worst outcomes may be dropped, and some costs ignored in cost-benefit accountings. In addition, experts like doctors are often given credit for solving problems that usually get better all by themselves with time.

References (17%) – References are customer testimonials selected by experts to show to potential customers. For experts who have many customers, it is too easy to select good-looking testimonials even if most customers get bad outcomes. But references mean more for experts who have very few customers.

Office, Person Neat Orderly (16%) – All else equal, it can be harder to deal with an expert who is disorganized and unprofessional. They might compensate with other advantages, but it makes sense for customers to prefer neat orderly experts. Which pushes experts to be neat and orderly.

Small Trial (15%) – This is the second item on this list related to if customers gain value from experts, aside from in extreme failures. This requires that some customer needs can be met quickly and cheaply, that customers can soon see if they’ve gained value in those cases, and that the value gained from larger longer-term projects be correlated to those of small shorter test projects. Experts gain incentives here to do well on small cheap fast projects, but not necessarily on the longer harder-to-evaluate projects.

Response Time (15%) – All else equal, customers would rather not wait long for an expert to respond to their phone calls or emails, wait long to get an appointment, or wait long once they’ve showed up on time for an appointment. This pushes experts to respond promptly to customers and to meet them on time. Except sometimes, as with cash prices, customers take a high waiting time price as a sign of high expert quality.

Fee Rule (13%) – Depending on the kind of expert, it can be a better sign about expert incentives if they charge per hour, or per problem, or in some other way. Customers who know the best way will pick experts who charge that way, which pushes experts toward that kind of fee rule. However, it can often be more profitable to pick the bad rule for ignorant customers.

Warrentee, Insurance (9%) – Experts who offer money back guarantees or insurance against big bad outcomes can reassure costumers regarding very bad outcomes. Like an expert quitting in the middle of a project, or taking your money but doing nothing. However, such insurance must be tied to very simple measures of customer outcomes, and money back guarantees only work if customers are wiling to give up on an expert entirely to switch to another one. These clues encourage experts to avoid these extreme bad outcomes, but don’t do much to encourage customer value in less extreme situations.

Rating Agency (9%) – Most endorsements are binary, or hard to compare with other endorsements. But ratings agencies often give non-binary ratings, on some sort of standard scale. This gives more info to customers, and thus more assurance that a particular expert is actually better than average. This encourages experts to coordinate to create ratings agencies that rate them highly. But it creates more conflict among experts of a given type for those higher ratings. Though often rating agencies try to pretend that almost all their experts are way above average.

Online Reviews By Customers (8%) – This isn’t as useful as customer referrals, but it is better than references, since these aren’t supposed to be selected by experts. However, experts often pay to get customers to post favorable reviews, or threaten to sue to discourage unfavorable reviews.

Size Of Organization (8%) – For some kinds of experts, their being part of a small or large organization can be a good or bad sign about their abilities to help customers. Customers who know this choose the better sizes, which can push experts to also pick appropriate organization sizes.

Match Demographics (7%) – All else equal it is often easier to deal with an expert who matches your demographics. Such as women dealing with women, or old experts dealing with old customers. This is an easy way to increase comfort, discussed above.

Fiduciary Standard (6%) – Experts who are legally obligated to follow a fiduciary standard are easier to sue if it can be later proved that they acted in ways contrary to their customer’s interest. As with insurance, while this can help to cut out some of the most extreme scenarios of poor service, it does little to encourage customer value in less extreme situations.

Portfolio Of Prior Work (5%) – This is the third item on this list that helps customers see value in non-extreme situations. Relative to a small trial, this allows customers to see value achieved in larger projects. However, this requires that customers be able to judge the value that other customers achieved in situations different from the one they now face. This encourages experts to serve each customer in ways that are easier to show and explain to other customers.

If we judge by the clues given in “How to Pick X” web guides, customers rarely attend to clues (Track Record, Small Trial, Portfolio Of Prior Work) that most directly show the value that customers achieve in typical situations. However, a few more clues (Disciplinary Actions, Complaints, Crimes, Warrentee, Insurance, Fiduciary Standard) also help customers avoid some extreme failures.

The most common clues attend to social proof (Credential, License, Customer Referrals, Experience, Location, Expert Referrals, Prof. Associations, Better Business Bureau, References, Rating Agency, Online Reviews By Customers). Almost as common are clues that indicate customer convenience (Price, Location, Office Person Neat Orderly, Online Listings, Match Demographics, Response Time).

The few remaining clues are commonly accepted if questionable indications of quality (Gut Comfort, School), and clues that would indicate higher customer value if there was any customer value to be had (Specialization, Experience, Fee Rule, Size Of Organization).

Actually, the situation is probably worse than this, as it is probably the most responsible customers who consult “How to Pick X” web guides, and look for the clues they mention. Most customers probably look at even less.

It seems that most customers of experts do little to check on if their experts are likely to actually give them substantial value, though customers do a bit more to protect themselves against the worse exploitation scenarios. Customers instead mostly seek social proof, convenience, and a few weak indicators of quality, assuming there is any quality to be had.