Dharmesh Shah wrote a great post yesterday entitled “Web 2.0 Revenues: Charge early, charge often.” If you’re not subscribed to his OnStartups blog, you should be.
In this post, Dharmesh essentially rebuts Paul Graham’s theory that you simply have to “build something people want” and can figure out how to monetize it later. He makes the compelling argument that you should use the “release early and often” approach with both your product and your business model.
I would argue that you can really sum all of this up by slightly modifying Paul’s basic premise: “Build something people will pay for.”
Clearly, this is a rule of thumb. YouTube is free, and people really want it. In order to break even, they’ll soon need to charge $100.00 per year per user. There are certainly exceptions to this rule, but these companies are rarely profitable. They can be acquired for huge numbers, and it’s the stuff of dreams. But if you’re not going to build a top 100 web site, this is not going to happen to you.
Can you offer a counterexample? Can you point to a purely web-based startup (i.e. no shipping dogfood) that became very popular, and yet whose owners didn’t ultimately figure out a way to make money from it? Excluding edge cases like startups that got sued out of existence, or imploded because of arguments between the founders, etc.
Hi Paul. Any company that becomes “very popular” (i.e. a top 100 site) is going to be able to make money (through ads, a strategic acquisition, etc). I agree. In fact, at the end of the post, I talked about this very point.
It’s completely reasonable to say “build something twenty million people want” without testing your business model up front.
So the difference in your statement and mine may simply a matter of context and the goals of the founder you are speaking to at the time. For example, if you’re saying this to someone who is building a vertical software solution with the potential to build a $20m/year business, I think you have to bring the business and revenue model up early in the discussion.
One of my favorite examples is Photobucket. They became profitable early, because they tested their business model and made sure that premium services plus ads equaled success. They didn’t wait to have 50 million users to start figuring this out. Had this not worked profitably they would be more like YouTube in terms of bleeding money. Both are likely to be ultimately successful, and perhaps already are depending on your definition.
I admire the work you are doing at Y Combinator. Thanks for the comment.
The definition of success is important. I think there is a Web 2.0 blinspot to the lessons of the dot com bust. I call it the three curses of Internet success: Success Without Revenue, Success Without Profits and Success Without Barriers to Entry. You may want to look at it on my blog: http://bookmansbusiness.blogspot.com