mokoyfman.com

The Myth of Free Classifieds

My friend Jeremy Philips wrote a great review for the WSJ recently of The Curse of the Mogul, a new book on the plight of the media industry by Jonathan Knee, Bruce Greenwald and Ava Seave.  I recently started reading the book, and it’s quite a good read.  But I want to focus here on one very insightful point Jeremy made in his piece:

For an example one need look no further than online classified advertising—which, the authors say, was the “first killer moneymaking application” on the Web. Leading online players around the world, charging fees, have withstood challenges from rivals offering listings free—suggesting significant competitive advantage. Craigslist, mythology aside, has been charging for job listings in its home market of San Francisco for more than a decade.

With all the (deserved) hoopla around the ”free” craigslist and its tremendous success, it’s important to remember that craigslist actually charges for key categories (e.g. jobs, real estate) in big markets (e.g. SF, NY).  Not only is this where they make their money, but its also how they make the site useful.  Without separating the wheat from the chaff in these key markets, the service would likely be overrun by spam and unusable for consumers.

This reality is consistent across all classified category leaders on the web. Leaders in Jobs (Monster, CareerBuilder), Personals (Match), Autos (AutoTrader), Real Estate (HomeAway) all have paid listings model that drive their business.

There are current attempts to buck this model, most notably Zillow in real estate and OLX in international markets.

Early on in the life cycle of a business such as Zillow, having a free platform is key to amassing listings and eyeballs.  I’m very curious to see how their model evolves over time though, as they continue to drive audience and scale of listings.

OLX doesn’t suffer from some of these issues in their smaller markets with low supply volume.  And it is my understanding that they are moving towards a paid listings model for exposure in key scale markets.

The only solution to this problem as I see it is a technical one that can separate the relevant listings from the noise for each unique end consumer.  This is seemingly a very challenging data normalization and search/algorithmic (or potentially social data/behavioral) problem that I haven’t seen a great solution to yet.  If you are out there cooking one up though, I’d love to hear about it…