When push becomes shove


By Doc Searls
March 9, 1997

"Throw away your browsers," says the breathless cover story in the May issue of WIRED. "Push technology is here!" It'll be like TV, only better! Hold still for "ambient broadcasting!"

Over in the February 24 issue of Business Week, the cover story about "Webcasting" reports that the Web is "a victim of its success" because "noise and congestion is making it hard for Web sites to attract visitors -- and keep them coming back." Which is tragic because without traffic these unhappy sites can't sell advertising, which "remains the most promising way to make money."

The CEO of a hot new "push" software company says "People want their computers to be as easy as their television..." And, "They just want a few channels they can turn to."

Just a few channels? Do people want "just a few books" in their libraries? Or even to limit to "few" the number of channels their cable brings in?

What kind of mass hypnosis is going on here? "The net for the first time is causing information overload," says Marc Andreessen. "Manually searching the web is not a sustainable model, long-term," says Eric Schmidt.

Do these guys mean that the only sustainable model is "pushing" data out to "viewers?" Do they really think all we want out of the Web is a narrow form of television?

Of course not. But that's what WIRED and Business Week want them to say, because they can't grok a medium where ads don't pay most of the bills.

Advertising today is a $160+ billion business*. Most of that money is spent by advertisers to "push" messages through media that offer no alternative. Those messages are "targeted" for "impact" that will "impress" a "brand" on the minds of those who, like cattle, hold still for the service.

Traditional broadcasting and publishing media are highly asymmetrical. They are one-way and one-to-many. The biggest media -- the ones that reach the largest of the many -- are extremely scarce. There are only four major TV networks, two or fewer daily newspapers for most cities, and a few dozen national print media that reach more than a single-digit wedge of the population.

The Internet is the best pull medium since the telephone.

The biggest advertisers pay an awful lot for ads in those media. The targets, however, pay nothing. They have no voluntary influence on the advertising business. Their job is to "consume" products and offer feedback in the form of cash. The "consumer" label testifies to this limited function.

This is why the market for advertising is not consumers but advertisers. It is also why the whole U.S. advertising industry is still a conversation between advertisers and media, not between either party and consumers.

Because consumers pay nothing for advertising, neither advertisers nor media get the message they would get if the relationship were symmetrical: there is no demand for mass media advertising. If anything, there is negative demand: people hate it.

This is not to say that all advertising is pushy, or that none of it respects the pull that makes a marketplace. There is a high degree of pull for advertising in trade publications, where the advertising often amounts to paid editorial content. And some forms of advertising -- notably newspaper classifieds and yellow page directories -- exist to satisfy customer demand for hard information just prior to making a purchase. This chart shows how the media vary in the level of user demand for their advertising:
Directional Facilitation
Demand for advertising
Yellow Pages
Direct Mail
none, unless solicited
only for shopping channels
Mass market magazines
low to moderate
Trade & special interest magazines
moderate to high (for classifieds)
low to moderate
high for practical info, low for slogans

The key is solicitation. Users don't ask for Web site banners, any more than drivers ask for billboards. They do, however, ask for advertising in specialty publications like Linux Journal. We can tell they ask for it because they pay for the product, the product includes advertising, and all three parties -- magazine, advertisers and readers -- belong to the Linux community and participate in the Linux movement.

The Internet, however, is essentially free. Customers may pay for access, much as they pay for cable television; but they don't expect advertising to show up as part of the bargain. And they certainly don't demand it. Or do they?

Fortunately, there is a way to test the value of advertising on the Web. Since the Internet is a symetrical two-way medium, it can facilitate negotiations over the values and prices of all kinds of things, including advertising . If it is true that users would rather reject than accept most advertising, the Web should yield negative advertising values.

Sure enough, this is how one company -- CyberGold -- is already starting to bet. CyberGold pays users to read advertisements. Nat Goldhaber, one of CyberGold's founders, says "Attention is becoming the scarce and valued commodity. If an advertiser would like to convey a message to a targeted customer then that customer ought to be compensated."

Needless to say, this is not the kind of thinking your traditional media types want to hear. After all, most of the media covering the Web are also part of the push marketing conversation. Naturally, they see the Web as a battleground for competing advertisers, and don't care much for non-consumptive pull-side behaviors.

Look for proof in the Web server business. Despite endless coverage of the "war" between Microsoft and Netscape, the top Web servers, by far, are from Apache, a "project" by a bunch of developers who work for free, or close to it. According to Netcraft, Apache has a 41.12% share of the Web server market, followed by NCSA, with 9.71%. Microsoft is third, and Netscape is fourth. Both have less than 10% shares. But try to find coverage of Apache and NCSA. They're nearly invisible, even though customers are pulling them into the marketplace at more than twice the rate of Microsoft and Netscape combined.

Esther Dyson says the big challenge today is not to add value but to subtract garbage. Most advertising is garbage. It's hard to imagine a less efficient way to communicate, or one that wastes more time and materials. Even direct mail, presumably one of the most personal and efficient forms of advertising, is so unwelcome and wasteful that its nickname -- junk mail -- is a synonym for garbage.

"Well, how else are we going to get our messages across?" the question goes. "And what else is going to pay for our publication (or radio station, or whatever)?"

The Internet finally gives us a chance to come up with imaginative answers to those questions -- answers that are not just another form of advertising. Maybe the new push technologies give us some of those answers. It is clear that most of them introduce useful efficiencies to the Web.** And most of them won't work without some kind of pull on the user's side.

At this point the problem isn't really with push technologies, but with the way they're being covered. The Internet is the best pull medium since the telephone. Treating it like a second wind for the overweight, government-maintained cancer patient TV has become is myopic and delusional.

There is plenty of money to be made through the Web (if not on it). That money will be made by companies who watch what the demand side pulls.

It won't be advertising. Or if it is, it won't look like most of the garbage that has gone by that name in the past.

* The latest numbers from McCann-Ericson are from December 1995, and are can be seen here, in a report that predicted a $161.5 billion 1996 advertising market in the U.S.

**For a sane overview of push technologies, read Web Informant #58 by David Strom. Also check his Push Publishing Technologies page, which has an excellent table of push companies and offerings, plus some other useful links.

To the top
To Previous Article
To Reality 2.0