This is something I struggled with for a long time in regards to my own company. I did ultimately though end up with going with a "Freemium" model though. The article does a good job pointing out the strengths of freemium, and I think for any sort of social network based company (I believe he used linkedin as an example) you would be hard pressed to find a pricing model that works as well as Freemium.
"Guess what? Because of declining hosting and bandwidth costs, for most Internet products the marginal cost today is practically … zero."
That's not really true at all. I remember reading an analysis of Google's costs somewhere (sorry tried but can't find the link) that each search costs something like 70c.
Even if it were much less (1c or 2c) that's still a hell of a cost for the "paying" customers (ie. the circa 10% of people that actually click on google ads) to be covering.
Yeah a trillion is a little high ... damn I wish I could find that article though it had this really good analysis of googles profits vs. number of searches and then inferred the cost per search. Mabye it was 0.7c :)
Agree. Hosting and bandwidth costs are just a part of the total costs of running an online service, besides development, customer support and marketing costs.
"In other words, if the cost to serve a customer (support aside) is zero, the long-term price of the product in the market will be zero (because of competitive pressure)."
I can't think of any large scale operation where the cost to serve a customer is zero. Sites like Google, Facebook and YouTube have massive operational expenses that they struggle to cover with aggressive advertising.
Of course, the freemium model works very well for many companies. Nevertheless - "zero marginal cost" is an illusion.
Although, to be fair, in the strictest sense of what he said, it still stands correct. None of the examples you have given hold up to the rule. All of the sites (Google, Facebook and YouTube) are not in a 'perfectly competitive' market due to network effects (even though Google and YouTube deny this).
If they were, none of them would own 60% of the total market. There would be many more market participants and there would be no barriers to entry. That's what he was referring to.
In a market where there are no barriers to entry (i.e. perfectly competitive), the long-term price will tend to zero.
Now that I am thinking about it, even though the internet is 'free', and has very low barriers to entry for some businesses, I can't think of an example of some sub-industries (search, social networking, video hosting, etc.) online where the market is perfectly competitive. There is always an inherent advantage and some outsized competitor, relative to everyone else.
"In a market where there are no barriers to entry (i.e. perfectly competitive), the long-term price will tend to zero."
Yes, the price will gravitate towards zero because of competition or due to the fact that end users would not be willing to pay for the product anyway (who would use Facebook if it wouldn't be free?). But the marginal cost of such operation would not be zero and costs would have to be covered by other sources (VC funding, advertising, premium services...)
You are spot on.
This false statement comes from the same misconception that started with "thanks the Cloud(tm)(r)(c) hosting my multimillion page views website is now free" that marketers spinned.
There has been a continuous price drop in commodities, it does not mean that the cost of operating these commodities drop at the same rate.
A whole operations team is needed to attend the infrastructure, whether or not your company actually owns its infrastructure.
Granted, thanks to automation, an operation team can attend to a larger number of servers than before, but this cancels out with the growth in the number of servers needed as the application complexity grows.
