Tales of Technology
A month ago I noticed a pain in my heel. So I went to Google and typed in "sore heel." I received a page of answers including its medical name, plantar fasciitis. There was an impressive treatise on the subject by an amateur. Most of the other answers I found were ads for devices like arch supports. The amateur said his favorite product was a particular kind of arch support, so I clicked on an ad, ordered it and had some in my shoes a few days later. Welcome to the world of Internet commerce!
Pre-Internet I might have called a doctor, but getting her attention for such a minor pain would have been inappropriate. More likely, I would have consulted a home medical book, then the yellow pages for an arch support store. In fact, I bought another pair of supports locally a few days after the first pair, but they don't seem as good and cost almost as much. In sum, the Internet took far less time and produced a better result.
Now let's look at this process from the seller's side. In the old days, the seller of arch supports would probably advertise in podiatrist's and runner's magazines, and put his product into shoe and sports stores. He could not have afforded to place an ad in a mass publication. Getting sales was expensive and hit-or-miss. Decisions about where to advertise and sell took many months to play out and gave inconclusive results.
In the Internet world the seller has new, much more powerful tools. The seller of arch supports contacts Google and says, "Whenever someone inquires about sore feet, arch supports, running, podiatry, or plantar fasciitis, please show them my short ad. For each person who clicks on my ad, I'll pay up to $2." Soon thereafter, the seller gets a lot of immediate information about interest in his product including how often inquiries lead to sales. His ads are shown to only those people who have already expressed an interest, and he doesn't have to pay unless they express further interest in his product by clicking.
Google is very sophisticated about how it shows ads. It doesn't simply show the ads of the highest bidders. It measures how often particular ads are clicked when shown in response to particular inquiries. Then, it shows the ads that are most likely to make money for Google. For example, suppose Google has noticed that one out of 100 searchers for "sore feet" end up clicking on A's ad and A offers $1 per click while only one out of 1,000 searchers click on B's ad which offers $5 per click. Google will show A's ad first because that maximizes Google's expected revenue.
This whole arrangement balances the interests of buyers, sellers and Google in an elegant way. Now, everyone's motivations are aligned. If I search for information about a problem, I'll be happy to see ads for solutions. The ads are unobtrusive and brief because Google wants me to click on them to see more. The sellers will be happy to pay Google if I click on their links, and Google will be happy to be paid.
Is there a flaw in this scheme? One problem is that clicks don't equal sales, and Google worries about people unleashing net-bots to click on rivals' ads as a way of making the rivals waste money. In principle, an Internet intermediary should be paid for its contribution to a sale. That, of course, is the business of eBay, Amazon and FreeMarkets.
I hate ads selling stuff I will never want, such as cat food -- at least until my retirement savings give out. The ads I like are the ones that tell me where to find something I really want, such as arch supports. In the middle are things I might want if I knew they existed, such as tickets to see Twyla Tharp. The problem with advertising is that it often misses its audience. Both the sponsors of the ads and the receivers have a common interest in avoiding misses. The Internet is offering to link sellers and buyers in a new way that will make many other kinds of ads obsolete.
Google takes advantage of the interactive nature of the Internet to serve the purpose of ads, linking buyers and sellers. Other people are using the Internet as if it were a just a one-way broadcast medium. They plaster their sites with ads, just as magazines do. If I go to a nonspecific site such as Yahoo, I'll see ads I have no interest in; today it is for "Pirates of the Caribbean."
Yahoo can claim that millions of people visit its front page, but the percentage of them who might be interested in a particular product is minuscule. Everyone else is just irritated by the wasted space. Some Internet ads are "improved" by making them flash and do other distracting things. A worse abuse is to make ads pop up in different windows that obscure the window one actually wanted to see. Not surprisingly, people hate these ads even more than the more familiar radio and TV ads.
Speaking of TV, TiVo shows that the days of conventional ads on TV are numbered. Once material is stored digitally on a disk, it is easy to skip commercials. I was able to hack my TiVo remote to skip forward in 30 second steps, and I haven't seen a commercial or Steelers' huddle in the past year. The only reason TiVo doesn't eliminate ads automatically is that it would drive the established channels crazy, and they would find a way to kill TiVo.
Someday you'll be talking to your TV system the way you talk to Amazon.com, the Internet bookseller. The TV will say, "Since you're watching 'Law and Order,' you might want to watch 'CSI'; 72 percent of Law and Order viewers also like CSI."
The technology behind this idea is a lot of cross-correlations computed on all subscribers and preferences. You will be able to conceal your preferences if you wish, but most people will happily reveal them since it may produce helpful advice. As the amount of data on preferences grows the advice gets better. Besides recommending TV programs, your TV might say, "Consider going to Duquesne Law School."
The Internet is a new communications medium for carrying out commerce. Commerce still needs middle men, but their role is going to drastically change in this new medium. And, yes, thank you, my heel is feeling better.