« Is There Such a Thing as Bigotry? | Main | The Butler Did It, of Course! »

January 24, 2007


Buyers would normally wish to associate with inferior, not superior, sellers. Suppose you had a choice of which dealership to visit to look at a car you're interested in but not sure is best for you. One dealership has superstar sellers who get almost everybody who comes in the door to buy a car at full price. The other dealership has mediocre sellers who get a much lower proportion of visitors to buy and often have to offer discounts off list price. Which dealership will you go to?

I think your explanation of advertizing is too specific, but that strange preferences are an important part of it.

Your comparison between buyers and sellers seems pretty off. We hear about sellers because of specialization. That may explain the publications, but don't forget Consumer Reports. I think rent-seeking is also worth mentioning.

Curt Adams: My impression is that most people would go with the first. You've loaded the question quite a bit, but I think only slight changes to what you've written, changes that would keep you believing that the second is the obvious answer, would produce an interesting survey.

Douglas, yes sellers are more specialized that buyers, but that is a puzzling phenomena in need of explanation.

Curt, holding constant the quality of the product, and if you only cared about quality and price, of course you'd want to frequent sellers who offered lower prices. The question is how reasonable those assumptions are about real sales.

I suspect the phrase "impressive seller" is not well defined here. Certainly the car seller that can get almost anyone on the lot to buy at full price will impress _other people in sales_, but could very well drive away purchasers who were aware of the seller's track record. On the other hand, many "big box" stores are impressive in terms of their scale and logistics capability.

This is the sort of claim for which there would have to be really, really robust empirical evidence to make it plausible.

You can impute all sorts of wacky preferences to buyers to make ad response rational. Maybe buyers like to purchase from sellers with good taste in music? Maybe they like to buy from the people whose ads support their favorite sports team? Maybe they like television so much that they'll buy products advertised on television just to ensure their favorite shows don't get cancelled? Maybe information costs about products are so massively significant to them compared to the differential benefits of various products that they will either purchase based on the di minimis amount of free information in advertising or choose products at random?

The common feature of all those stories is that they defy our self-understanding of what we do when we choose to participate in a transaction, and thus demand evidence. Unless there's some reason to believe buyers actually have the preferences in question, what's the point?

Of course it depends on the definition of "impressive salesperson". But I think the normal definition is that somebody who can conclude deals more often is a impressive salesperson. Since there's almost inevitably negotiation involved in any sales transaction, a impressive salesperson will normally be able to extract a better price as well. So I think it's very reasonable to say if you buy from a impressive salesperson you will more often be convinced to buy and often pay a higher price. Buying more often isn't always bad, but I think in the modern world it usually is. If you know the product is super-valuable you don't need a salesperson anyway. Paying more is of course essentially always bad. So smart buyers associate with poor sellers. The only substantial exception would be if the salesperson is highly informative - which is 4) on the existing list. I know when I consider dealing with good information-based salesperson I consider their high-quality sales ability, and ability to extract more money from me, a *detriment* I may choose to put up with for the benefit of the informational product they can provide. So even here, insofar as 5) is in play, it's a detriment.

Robin, why is the *general* tendency of salespeople to be more specialized puzzling? In the car example, I'll buy a car once every 5 years, and the salesman will sell several every week. (I know that's relatively extreme but that's the general pattern.) When the buying is frequent, as in company buyers, buyers *are* often quite specialized and very good at what they do. Ask anybody who's sold to Walmart. But the consumer market is basically always going to have more specialized sellers, and since that's 2/3 of the economy in general sellers will be more specialized.

I think sellers are more valued also because, usually, more gain is possible by being a super-excellent seller than a super-excellent buyer. A super-excellent seller can theoretically take over an entire market, at least to the point of spending all his working hours writing highly profitable contracts. Consumer buyers, and to some extent commercial buyers, are limited by their demand. If sales and buying ability are completely transferable then the imbalance probably goes away for commercial buyers, but will remain for consumers.

Curt, there is ambiguity in what it means to be specialized if we are thinking about knowledge. The car seller knows more about cars than you, but you know more about your needs than he does. We each sell our labor to companies which buy it, and we may know more about our abilities and they know more about their needs.

So perhaps it is better to define "specialized" in terms of a frequency of similar transactions. Company buyers of our labor are in this sense more specialized than we are, and I'll agree they are relatively expert. My point is that such buyers don't have near as large a place in our imagination as sellers.

Paul, as I see it my theory fits lots of otherwise puzzling data, and any theory of the effectiveness of uninformative ads is going to be somewhat at odds with our self-understanding. But I'm open to hearing other suggestions. There must be some explanation.

Robin, your (1) "exploiting bias" lumps together a huge amount of territory - all the biases that ads could conceivably exploit - social proof, exposure effect, mere recognition, halo effect, anchoring on "buy 12 for $24", and good-old-fashioned Pavlovian association to the hot girl lounging on the hood of the car.

Isn't this the customary, standard explanation for the effectiveness of advertising among social psychologists? I just don't see what is the unexplained surprise here. The cognitive biases seem to cover the surprise that advertising is effective; and the asymmetry of incentive seems to cover the surprise that advertising is more talked-about than buying (except among professional corporate buyers, who buy particular products frequently enough to make worthwhile acquiring specialized knowledge in them).

Is a critic a kind of specialized buyer?

Also, perhaps "super sellers" - those that are more effective at getting people to buy products for more money - actually add value for the buyer by finding the buyer a product that better fits their preferences, even if they might end up paying more than they otherwise would have? If I knew exactly what car I wanted, all I'd have to do is look for the "best deal." If I knew that I wanted a car but didn't know which one, a "super salesman" might be more useful than a poor one.

Eliezer, social psychologists may prefer to explain ad effectiveness in terms of the biases that they study, but economists are not so sure. And it is not clear incentives are asymmetric.

The comments to this entry are closed.

Less Wrong (sister site)

May 2009

Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30