msdewey(Note: this article discusses pornography but has no X-rated links and should be safe in PG environments. School teachers and clergy are duly warned.)

If the web has let a thousand flowers bloom, pornography has been its most prolific deflowerer. One study found a third of Internet users visit an adult site at least once a month (a number that sounds low to me), and one in four workers view porn during work hours, the most frequent time of day for visits. The $100 billion pornography market is a fountainhead of innovation and the most lucrative type of content: an estimated $3,000 is spent every second on the web’s 420 million adult pages (12% of the web). This is a tail longer than Ron Jeremy.

With so much web porn available, how do people decide where to get it? The most common search queries used to be porn-related, but navigational queries are now more popular, suggesting searchers are going directly to their favorite sources (social networks may also be co-opting adult traffic, suggesting people are focusing on the real McCoy). Aggregators like Xvideos, RedTube, and YouPorn have made finding porn easier, but these sites do a poor job of personalization.

In a category where people have such deep and varied tastes, there is an opportunity to create a StumbleUpon for porn. Let’s call this hypothetical product Frisky.

Frisky could be a toolbar like personalized recommendation service StumbleUpon but would likely get traction faster as a website like Digg. Users would sign up, enter demographics, and select their preferences for many criteria: orientation, age, race, fetish, number of participants, media type (photos/videos/audio/slash fiction), and more. Users might select specific stars they like and categories they don’t like, since repulsive content can kill interest faster than a mullet.

Users could then click, rate, and comment on recommendations, making Frisky continually smarter about the tastes of each user and the community. Crawling content would not be hard — the web won’t run out of porn any time soon — and most adult sites would welcome the incoming traffic; they would probably submit a lot of content themselves. Frisky would not host content but merely link to it, avoiding hosting costs and conflicts of interest with content sites.

While a lot of adult content has metadata like orientation and category, Frisky could solicit even more tags from users. This would allow further personalization, flag spam, and create valuable metadata to share with partners, who could then enhance their own tags.

Frisky’s main revenue model would be lucrative affiliate fees from paid subscription and DVD sites, which typically pay $30-50 per subscription or 40-50% revenue shares. Frisky could also sell reports on consumption patterns, with likely valuable insights since research in this area is challenging.

Marketing may be a challenge since porn is a private, anti-viral application. One strategy would be Digg-like widgets for content sites seeking more incoming traffic. Search engine marketing and partner advertising are also likely feasible given Frisky’s high monetization potential. Porn stars and companies could also create site profiles to market their popular content and interact with fans.

Frisky’s deep personalization would be a significant competitive advantage over adult search engines like Booble and aggregators like RedTube, most of which have only general popularity ratings. Frisky would also benefit from a slight network effect, since its engine would become continually smarter with every new user.

Other competitors include StumbleUpon itself, which has a porn category but doesn’t have deep categorization or targeted features. This is the same reason Pandora is a better source for popular music. Many of StumbleUpon’s top listings also just go to porn aggregators instead of specific content. StumblePorn was a Firefox install that didn’t get traction and desperately sought a small sale. AdultStumble also came and went, so to speak.

The porn market is not appealing to a lot of entrepreneurs, and it’s not one I’d personally try, but an intrepid team could build this quickly. Long tail markets are ripe for discovery engines and few tastes are more insatiable. Executed well, Frisky could bring a new level of spiciness to serendipity.

What do you think?

<br /> <a href=”http://answers.polldaddy.com/poll/1756172/” mce_href=”http://answers.polldaddy.com/poll/1756172/”>Do you think Frisky is a good startup idea?</a><span style=”font-size:9px;” mce_style=”font-size:9px;”>(<a href=”http://www.polldaddy.com” mce_href=”http://www.polldaddy.com”>surveys</a>)</span><br />

[Photo credit: http://www.flickr.com/photos/dannysullivan/292257748/]

Want to see previous columns?
What’s Next: a Mint for location?
What’s Next: a Pandora for fashion?
What’s Next: Free computers for small businesses?
10 lessons from a failed startup

goldensonMark Goldenson wonders if anyone has ever used Chrome’s incognito mode for non-porn purposes (birthday gifts? riiight). He’s starting an innovative venture in health care. To submit an idea for the What’s Next series, email Mark at mjgold3@gmail.com. Selected ideas will receive attribution.