SuitWatchFriday, November 30, 2001
A Linux model for advertising
The advertising business is in a world of hurt. It started to get bad with the dot-com crash, since a huge hunk of that VC money (which peaked at over $25 billion/year for Silicon Valley alone in early 2000) was flushed right into "branding" -- which, common dumbness insists, requires massive amounts of advertising. Then things really went to hell after 9/11.
The latest episode in this tale of woe was told in Wednesday's Wall Street Journal by Vanessa O'Connell in a piece titled, "Decline in Advertising Revenue Worsens, Suggesting No Fast Turnaround in Sight". Here's the operative paragraph:
"A certain amount of paralysis still exists," said Lauren Rich Fine, a Merrill Lynch analyst who follows advertising companies. Ms. Fine blames a "9/11 hangover" for the continued drop in ad spending. Although demand for network-television time appears to be improving, she points out that marketers still aren't promoting their brands at levels close to those of the summer. (The piece is here, but you may need to register.)
Newspaper advertising was down 11.5% in October, Fine reports, after a September drop of 9.5%. TV advertising declines were projected at 15% to 20% in November and December, against the same period one year earlier. Radio and magazines were nearly as bad.
But the overall trends were down already. Bob Coen of McCann-Erickson, the eminent analyst of advertising trends within the industry, had already published a glum outlook early this year, blaming the dot-com bust for much of it.
I bring all this up because I know more than an editor is supposed to know about advertising and marketing, and I don't want it to go to waste. I also see a larger story going on here -- one that's especially meaningful for Linux and the people who put it to use.
Some background. Before Phil Hughes (Linux Journal's founder and publisher) talked me into going back into full-time journalism several years ago, I was a freelance writer making most of my money as a marketing consultant. Before that I ran the PR agency that spun off Hodskins Simone & Searls, the advertising agency I co-founded in 1978, and which became one of the top high-tech agencies in Silicon Valley in the late 80s and early 90s (it was acquired by Publicis in 1998, long after I had left). And before all that I was a journalist. That's what they call a "career arc".
But I was a long shot from your average marketing dweeb. While others in the trade worked as corporate cosmeticians, I billed myself as a hired incendiary. To filter out potentially boring clients, I presented my marketing logic this way:
I submit Linux and the methods that produced it (courtesy of the free software and open source movements) as Exhibit A in my case for Marketing as Arson. Exhibit B is The Cluetrain Manifesto, a Web site and book I co-wrote about "the end of business as usual" (to quote the subtitle). Cluetrian's marketing budget rounded to zero. There was no "branding" involved, and the book became a best-seller.
Nor was there much marketing for Linux, except during the dot-com boom, which only leveraged Linux' already-proven success. And today Linux is such a big "brand" that Microsoft continues to position it as that company's "number one threat" (to quote Steve Ballmer).
So what's a "brand" anyway?
"Branding" is an idea Procter & Gamble got from the cattle industry around 1930, when it discovered it was possible to expand shelf space in grocery stores by putting one kind of soap in six different boxes and singing about the difference on the radio -- and doing it so aggressively that the sponsored programs became known as "soap operas".
The era of Mass Media ran roughly from the rise of commercial broadcasting to the rise of noncommercial networking, which produced the Internet. No matter how commercial the Internet becomes, nothing can alter the power balance it restored to markets. Thanks to the Net, supply and demand -- vendors and customers -- are one link apart. Markets are truly conversational, for the first time since Industry won the Industrial Revolution. And the supply side is no longer in charge, which is what you've got to have if you're going to do mass marketing.
Linux is a product of conversational marketing. It branded itself. Today Google finds over 41 million pages on the web that mention Linux. That's about four times as many as those that mention "camera". Not bad.
So what becomes of advertising in the era of conversational marketing? I think it's going to be fine. But it will have to make as much sense to those who get it as to those that give it, and this is an alien concept to the advertising business.
Case in point. For decades the advertising business has been obsessed with creating a single product: the message.
Right now thousands -- perhaps millions -- of marketing, advertising and PR folks are laboring to produce a "message" of some kind. Yet it's a rare customer who opens a magazine or turns on a TV looking for "a message". One of the persistently undetected ironies of our time (and the past seven decades) is that there is no demand for messages. Just like there is no demand for "content." Nobody goes to a Web page saying to themselves, "I wonder what content wubble.com is delivering today". Both "message" and "content" are abstractions born of detachment from reality on the demand side of the marketplace. They are conceits of supply. And they don't work in a world where markets are fully conversational.
So how can advertising make sense? Easy: by being efficient on the supply side and useful on the demand side. It isn't complicated (though your advertising agency will be glad to complicate it for you).
Being efficient and useful is easier for some media than for others.
Generally speaking, the more specific a medium's subject, the more useful it is for its audience and the more efficient it is for its advertisers. Consider a trade publication like Linux Journal. The advertising it contains makes it more valuable, not less. Same is true for the classified ads in newspapers. In the mass-marketing world where commercial network TV lives, the problem can get really tough. Your MUTE button is there for a reason. It testifies to the negative value of most advertising messages. That's also why people pay extra to HBO and their local PBS stations.
You'd think, then, that the most useful and efficient advertising medium would be the Web. The problem is, users are less inclined to perceive the Web as a "medium" than a place or a space -- and a rather private one at that (kind of like the telephone, where we've never welcomed any advertising). So the challenge for advertisers is to make themselves both useful and welcome.
I believe this calls for an approach to advertising that comes from the same unexpected angle both Linux and the Internet came from. It won't be easy. But it's hard to find two better models for usefulness and efficiency?