Esther Dyson is Selling Something

A long time ago, in a galaxy far, far away, Esther Dyson was a journalist.

Now she’s a venture capitalist, but most people still think she should be accorded a journalist’s credibility.

That is wrong.

When she speaks now, she is usually selling something. She’s a saleswoman. Not that there’s anything wrong with that, but when a salesman talks to you a light flashes on-and-off for you, shouting "salesman"! Journalists are expected to tell you the truth, without fear or favor.

And right now she’s selling the idea of behavior being more important than content in building value for Web ads

Salesmen should not be writing what purport to be journalism columns, but I haven’t confused The Wall Street Journal with journalism for some time, so there you are.

Fact is we went through this dance about behavioral or contextual Web advertising models a decade ago. Among those trying to build contextual models was Engage, a unit of CMGI which was one of the most notorious flame-outs of the dot-bomb era.

The plain fact was that it cost real money to create behavioral models
for Web advertising, and you never got the results. The difference in
clickthroughs between an ad placed based on content and one based on
the user’s behavior was almost nothing. And, it turned out, the ads
based on content actually delivered more revenue.

If you think about this for even a moment you’ll see why this is so.

Your behaviors, the lifestyle which informs your purchase choices, is
usually at the back of your mind. The content you’re looking at is at
the front of your mind. It’s what you clicked to see. Which signal are
you most likely to obey if an ad pops up right now — the impulse at
the back of your mind or the one which directed you to the content?

Exactly.

But this doesn’t stop people like Esther Dyson. As the note at the bottom of her WSJ essay notes,
she’s got a ton of investments in behavioral-based advertising, and in
ad agencies in general. There are a lot of otherwise-smart people who
seem to think the behavioralists are finally going to win out, thanks
to social networking.

And in a way, maybe that’s true. If you click to a page run by someone
you trust, someone you agree with, and that page is recommending
something you’re behaviorally tuned to want, you might just go over and
buy it. But what actually happened? First, you clicked to the page of
someone you trusted. The content is the page of the person, not the
behavior. You first moved the impulse to the front of your mind by
clicking to the page of your friend, and then acted on it.

Now scammers may say, ah, if I just become a friend of a lot of people
they’ll buy what I say, and then I’ll sell my soul to advertisers and
clean up. But you’re confusing acquaintance with friendship. Just
because I’ve got an indirect link to you via a social networking site
doesn’t mean I’m your Best Friend Forever. Becoming a BFF requires a lot
more than a casual online acquaintance. I actually have to get to know
you, and you me, and we have to bond. Which takes a lot of time. And
effort. Which defeats the whole purpose the scammer had in mind — they
just want a ton of sales, not a complex web of real relationships.

It’s like the difference between a pick-up at a bar and a marriage.

Note that in all these examples what we think of as content has
changed, a bit. You’re not looking here to a page put up by a publisher
for a particular story on a specific subject, say Web advertising.
You’re looking to see what Dana thinks. Or you’re coming through a site
like Digg, through knowledge that this is about Web advertising, and
you have a mild interest in what the author thinks.

If you really know me, if you really trust me, you’ll buy my recommendation now like magic.
If you really don’t know me, if you came here via a second-or-third
degree link, whether based on context or behavior, you didn’t click
that link above. And you certainly didn’t buy.

That’s the difference.