For a couple of years here, I've been working on an idea that I like to call The End of Advertising As We Know It. Basically, it's where I see people like Google headed, and I think we can offer a better experience.

THE PROBLEM
Advertising is annoying, but in a very specific way: it annoys people only when it's poorly targeted, and that's because it's wasting time. And currently in this world, the vast majority of advertising is poorly targeted. I don't have any real numbers here, but I'd guess that less than 5% (perhaps way less) of advertising reaches its intended target -- those who will either remember the advertised message or actually act positively on it.

(Please keep in mind that I'm talking about all forms of advertising at this point -- not just ads you see on the web.)

The problem isn't the advertisements themselves. The problem is a nonspecific distribution system. Currently, targeting is very crude -- and it must be. This is because fine levels of detail about each potential recipient are out of reach of the networks that distribute the ad messages. It's not that the people that create and publish (and these are two different groups, by the way) truly desire to annoy the 95% of people that see their ads and feel no relevance to them. That wastefulness costs everyone -- it costs time, energy, and money. Nevermind the aggravation. But it's worth it to hit the five people in a hundred that think, "That is just what I needed, right now! Where do I sign?"

In many ways, the industry is still working in the days of basic statistics: creating general theories from patterns of data sampled from very small sections of a larger population. If you could go back in time and as Gallup about his Poll, or Nielsen about his Ratings, I bet you'd hear them both say, "Well, whippersnapper, my work is only an estimate based on asking a few thousand people out of millions. It's a good way to guess at something, because it's impossible to know people's preferences, short of storing all that information and asking them every once in a while. Statistics, man! it's the future! Because there will never be a warehouse on Earth that could store all of those punchcards!"


A SOLUTION
Well, it's the 21st century and it turns out that not only do we have the warehouse, we have the data terminals in front of just about every recipient. I speak of course of The Internet. How can we make it useful? How could we solve the problem of poor targeting, and make it so that each person likes every (or nearly every) advertisement that they see?

Let's take a look at what would make for good targeting. I believe that if you wanted to target the majority of advertisements successfully, you'd need to know at least the following:

1. Where the recipient is generally physically located
2. Basic stats about the recipient (gender, age, etc.)
3. What the recipient is interested in (hobbies, etc.)
4. The recipient's tastes in clothes, music, food, and other like items
5. Perhaps places and things the recipient already enjoys (and maybe even dislikes, so you can steer away from ads about those)
6. Whether or not the recipient is looking to solve certain problems (because after all, each ad is a possible solution to a problem)

With all that information about each person, and a sufficiently characterized inventory of ads, the prospect of specifically targeted ad messages becomes promising. That college kid who has 6GB of electronic music MP3s and likes to travel? You can tell him about the Winter Music Festival in Miami. The stay-at-home mom with serious craft skills and more spare time, now that the kids are at school? You can tell her about the online market for hand-knit goods.

Astute readers may notice that the above numbered list fits five major categories, and that each of these categories of information are already available on Tribe:

- Geographical relevance (1): our local focus
- Biographical information (2,4): user profiles
- Affinities (3,5): tribes
- Ratings (5): reviews and recommendations
- Needs (6): listings and requests

With these five categories, it becomes possible to put advertising in proper context, on a per-person basis, based on as much information as each individual is willing to share. We'll start out with your location (which is pretty much the baseline for non-member users) and then build from there. If you become a member, we'll take what we know about you (depending on what you're comfortable telling us) and show you some more items that you might find interesting.

I believe there is some external validation for all of this. In recent months, I have seen search engines and major online retailers build or purchase products that work along one or more of those five axes. I can only assume that they have the same goal: to better target their information, products or ads to potential consumers.

And it is the way things have always been: as buyers become more sophisticated in their awareness of markets and products, sellers are going to have to find better ways to get their messages across. The next step in the evolution of commerce will involve the refinement of this dialogue. More productive conversation to solve the buyer's problems, and fewer non sequitur, poorly-targeted messages.


DESIRED GOOD VS. NECESSARY EVIL
What I want to emphasize is that Tribe has never been intended to become a giant ad farm. Long-time users of Tribe will recall that we started out with no conventional display/banner advertising, and that when we started putting ads on pages, we explained that it was an experiment more than anything. We did not expect to make money from advertisements in the conventional sense.

For me, both of these sentiments still hold true. In the long run, we're not going to be in the business of interrupting content with the same conventional ads that everyone has to see. We are going to be in the business of placing information in the right place, for each specific user -- whether that information is a personal message, a group discussion, a listing for a used car that a friend is selling, or an ad for an online sale for flat-panel TVs.

What we've been doing here at Tribe for the last couple of years is working on this idea that when people come together and talk about things, transactions occur in a natural way. These transactions may not (and usually do not) involve money changing hands. But they do involve lots of good information. I believe another way to say this is that conversations generate leads. If we build a network that can distribute all items to the people that will be interested in them, then we'll have built an excellent system for distributing leads as well. And then the focus will turn to tuning the balance of ads/leads to make this a self-sustaining, and hopefully profitable business.

(When I say profitable, I mean it in several ways: profitable to users that save time and money, profitable to our bottom line as a company, and profitable by virtue of improving what I view to be a wasteful and irritating standard of distributing advertising vs. content. This is why I refer to this idea as The End of Advertising As We Know It: no more scattershot broadcasts of the same ads to the every person, with multiple impressions to press an unwanted message. Instead, fewer, more pleasing messages that are actually useful.)