I’m a guest lecturer for an executive MBA class at Denver University later today. I was asked to talk about Internet business models (among other things), so I thought I’d take a look at the 39 companies that have been through TechStars to give them a sense of the relative popularity of various business models. I think this represents a fairly decent cross section of reality, since about 75% of these companies have raised outside funding after TechStars ended.

Some of these are open to interpretation or are really a hybrid of a couple of forms. My categories might be somewhat arbitrary. But here’s the data as I see it:

SaaS (33%) – These companies sell their product to customers via the web, and don’t bother with a “try before you buy” product. Examples include Rezora, SendGrid and Filtrbox.

Freemium (20%) – These companies give away a free product, and then try to upsell more sophisticated features or solutions on top of that. Check out Baydin and TimZon.

Sell Installed App (5%) – I broke this out from SaaS, because these companies are actually selling software that is licensed and physically installed. Subtle difference these days. A good example is RedLaser from Occipital.

If you add the three approaches above together, you get 58%. So well more than half of the companies we’ve funded are ultimately selling software to people who pay for it. Novel idea, huh?

Gather/sell eyeballs (18%) – You might call this the “advertising” model, or the “underpants” model. Some call it “audience aggregation”. Some of the companies which achieved early exits were in this category (Socialthing, Intense Debate) , but it’s quite risky too. Often, companies doing this initially have other models in mind once they reach a critical mass but can’t use that approach early on because they don’t have enough scale.

Marketplace (13%) – These companies try to aggregate buyers and sellers, and generally take commissions or service fees for providing the marketplace. Foodzie and oneforty are examples.

Lead Gen (5%) – These companies often provide a valuable free service, and then provide qualified leads to buyers. This is very similar to the Freemium model, except that the upsell is not more software, it’s other services or software provided by someone else.

Virtual Goods (2%) – This model typically involves providing a game or other interesting virtual environment and then selling virtual goods in that environment. J-Squared Media’s MiniPlanet is a strong example.

Crowdsourcing (SaaS) (2%) – These companies use the power of a large distributed workforce, often to do things that computers can’t do automatically or efficiently. Typically they ultimately deliver a service to the customer. Retel Technologies is a good example.

Content Production (2%) – Although it’s a perennially unpopular approach with investors, these companies create content and then attract an audience for that content, typically selling advertising inventory targeted at the audience or subscriptions. Howard Lindzon’s WallStrip is an example of this approach that worked well.

Enterprise 2.0 (0%) – I was surprised to see that we haven’t funded any companies (yet) that are taking web 2.0 consumer technologies and applying them to enterprise settings. Some companies I know of that are doing these sorts of things are Yammer and Brainpark, as examples.

So there you have it. If you’d categorize the models differently, please let me know in the comments.