I’ve given this a little thought, and I’ve come up with six different activities people do on-line. These are only broad categories, and there is some overlap. But I think you can sum up what people do into the following six things:

(a) Reading Information. That is, you go on-line to find information: the current news, weather, sports scores, and the like. While this somewhat overlaps the “passive entertainment” category, other things in this category may include reading on-line computer programming manuals or reading mathematics tutorials.

(b) Searching. This somewhat overlaps the “reading information” category, but by “searching” I mean actively going to a search site such as Yahoo! search or Google search, and repeatedly iterating through a few keywords until you find something you are looking for.

(c) Interacting. This covers the entire gamut from reading LiveJournal to checking your e-mail on Yahoo! mail to texting people on Twitter. You’re talking to people or reading their responses. And while this does overlap with “reading information” and with both “passive entertainment” and “active entertainment”, the primary focus of this activity is interacting with other people.

(d) Buying Stuff. Amazon.com is a prime example of this sort of activity: you go on-line to buy a good or service or product which is either downloaded to your system or which is sent to you via mail.

(e) Passive Entertainment. Television is an excellent example of this, as is YouTube: you’re sitting in front of the computer being entertained. The flow of information is one-way, and while this can overlap activities such as “reading information”, “interacting” and “searching”, generally the point is to sit back and enjoy.

(f) Active Entertainment. By “active” I mean game playing, as opposed to watching YouTube: you’re online to play a game. And while this can overlap with “interacting”, active entertainment involves both a structured environment (an on-line game) and established goals. (I’d categorize Second Life as “interacting” rather than “game play”, as there is no structured goals within Second Life.)

Now companies like Yahoo! and Google and even on-line publishers make their money through advertising. And there are literally a ton of social web sites cropping up which are based on the idea that once you drive eyeballs to your site, you can somehow monetize those eyeballs through advertising. After all, that’s how Google made it’s money: by driving eyeballs to Google search, they’re making money hand over fist by running ads on Google search.

But…

But let’s think about on-line advertising in terms of the six activities above.

Interacting. When you’re interacting with friends, you’re focused on interacting with friends. Advertising–outside of a recommendation made to you by your friend, which could be a viral ad–simply won’t work. And it won’t work for a simple reason: the advertisement is orthogonal to–and distracting from–what you’re doing. Basically, to convert that eyeball to a sale that person has to stop interacting with his friends: a change of context from one activity to another.

Active Entertainment. Like interacting with friends, active entertainment involves actively participating in a particular activity–and unless what you’re trying to sell directly relates to that activity, such as advertising a sequel to the game, you’re essentially forcing the person to change what they’re doing from one activity to another. And while there is a lot of experimentation in in-game advertising, I suspect the return on investment will suck, because in-game advertising has the disadvantage of being like billboard advertising: we’re programmed to ignore it–but add that to the fact that it’s a path we’ll probably only go down a few times and then never travel again, after we’ve completed the game.

Reading Information. Web sites like CNN and Weather.com are attempting to make money through related advertising–something that Yahoo calls “context match”, where the ad is matched to the content of the web site. But if you’re on a site looking for information, chances are you either are engaged in “passive entertainment”–that is, you’re just reading for the heck of it–or you’re reading information. While I’ll cover “passive entertainment” later, the problem with reading information is that unless there is an ad which is directly related to what you’re reading about (such as a calculus textbook advertised on a web site covering calculus), the chances of converting the person’s information seeking activities into a sale strike me as slim to none.

Passive Entertainment. The advantage of advertising to someone who is engaged in passive entertainment is that if they’re just flipping through a bunch of pages, chances are they may click on your page if it looks interesting enough. So the click-through rate for advertising to someone engaged in passive entertainment may be fairly high. But the ROI will undoubtedly suck, as people engaged in passive entertainment are at best window-shopping–and it’s fairly unlikely they’ll bookmark your page to return on a later date, since your page is just one of hundreds being flipped through by someone who is just “browsing.”

Buying Stuff. Someone who is on your web site looking to buy something could possibly be upsold–so it may be a good idea to make the product your selling easy to categorize so they can easily discover additional products that they may want to buy. But unless you’re running a large web site like Amazon, you may not have a lot of stuff to upsell to begin with.

On the other hand, one problem I’ve experienced with Amazon is that related products are poorly categorized: when searching for a computer game for my Macintosh I may find myself on a page with the Windows version–yet no link to take me to the Mac version. And Amazon’s current algorithm for matching related products sucks, as it seems to simply weigh by what people buy rather than by what is related. (Meaning if I’m looking for a computer game, don’t try to sell me a DVD as well.) In this way I think companies doing on-line sales are leaving money on the table: Amazon, for example, is excellent at allowing me to find something I know I want, but lousy at showing me related things that I may want to buy.

Searching. Here is the cash cow of the Internet: if I’m looking for something, and a ‘sponsored link’ gives the result of what I want, and I’m looking for something to buy–well, chances are you’ll have an excellent ROI since I’m there looking to spend money.

I think Google and Yahoo’s Overture found the cash cow of the Internet: search. Search advertising is a cash cow, and it’s being milked quite nicely–but it’s a cash cow because advertising plays directly into the activity the user is already engaged in. Passive entertainment advertising may currently be a cash cow–but at some point advertisers may start realizing that the return on investment is just not there, since the real goal of most advertisers is not to gain eyeballs–but to sell stuff. And the conversion on passive advertising is really not all that high.

Web sites like Yahoo! Mail and MySpace and LiveJournal–advertising there is going to suck big time, as unless you’re doing “feel good” brand awareness ads, chances are, you’ll never get a conversion. And for web sites like Wikipedia, the only way advertising will work is if it is specific and specialized: a wiki link on statistics won’t sell a lot of windows games, but it may sell a lot of scientific calculators and statistics books.

Unfortunately, however, most Internet companies are still thinking of their users as undifferentiated eyeballs–and are operating under the delusion that those eyeballs can convert at the same rate. Which is why companies like Yahoo! are doomed: unless Yahoo! can figure out a better way to bolster its search advertising by driving more users to on-line Yahoo! search, they’ll never be able to monetize the eyeballs to the degree they think.