Tuesday, July 18, 2006

The Long Tail of Advertising

Have you ever written an advertisement? This blog is an ad of sorts, but I'm talking about a real honest-to-goodness ad. Something that tries to capture your imagination. In 100 words or less. In HTML. With little or no graphics. Sound daunting? It is. However, it's also only $30, so I can try, try again. Wecome to a revolution!

In the past someone like me would have to shell out a very uncomfortable sum of money to publish an ad in one of the 'recognized' industry periodicals. And I don't even dream about broadcasting like television. Instead, I dream of niche sites that have a greater proportion of interested parties and of ads that cost less to both purchase and produce. And I dream its so cheap I can set up metrics, produce a series and track the results over a one or two month campaign. It's how the big boys with multi-million dollar budgets do it, only I'm a small boy with a very small budget.

Luckily, sometimes dreams do come true. Perhaps you've heard the Long Tail of Merchandising, as espoused by Chris Anderson of Wired Magazine? Well, the converse is also true, and its called the Long Tail of Advertising. And it turns out Chris has a book on on the whole mishegoss. It's a definite must read, with the particular twist that I'll get to read about what I'm doing.

The beautiful thing is it brings cost of scale so that advertising and marketing campaigns are available for the rest of us. And you can argue its even more cost effective since the segments have a higher proportion of interested parties.

Which brings us to the new problem. I can't write ads. This blog is sort of an ad, but not really. Here I can talk at length-- which I have no problem doing. But an ad requires you to be brief, understandable and captivating. While I was an English major for a while, I did in fact end up with a Computer Science degree, and producing brief, understandable and captivating prose scares me.

Not advertising is not an option. So I have to start somewhere. Contrary to popular belief, merely placing a shingle on the web does not lead to your product being sold, tried or even known. It may be awkward at first, just like a homemade video uploaded to youtube, but it'll get better.

Why? Because as a computer scientist I will try, try again. That is, run many ads over several weeks with slight variations to content, and measuring the results of each run. I still have to figure out how to account for randomness-- things like vacations, placement, lead stories, moods, etc-- but I am a computer scientist and if there's one thing I can do is repeat myself again and again.

Which reminds me of a joke I heard a long time ago.

A physicist, a statistician and a computer scientist were staying at a hotel. Due to some rather unfortunate wiring, each experienced a small fire in their rooms while they were sleeping. The physicist sprang out of bed and bounded to his notepad where he performed a series of intense calculations over hydrodynamics and splash patterns. When he arrived at the answer, he filled a glass with the exact amount of water and threw it in the precise spot to douse the flames.

The statistician, upon waking up with flames in the room, leapt to his suitcase and found the specially designed faucet hose attachment he had custom made because he knew the odds of experiencing a fire in a hotel room were to great to ignore. Turning the water on full blast, he proceeded to fill his room with an inch of water in order to reduce the chance of further flames to nil.

When the computer scientist woke up with flames in the room, he opened his notebook computer, which was right beside him in bed. He wrote a quick graphical UI that asked the user to input the amount of fire, in joules, and the requested amount of water, in ounces. He then ran it a couple of times until he was satisfied, finished with a nifty animation of water flying across a room and smothering the fire, and went back to sleep.

No comments: