Hacker News new | comments | ask | show | jobs | submit login

"Although about 3 million computers get sold every year in China, people don't pay for the software. Someday they will, though. And as long as they're going to steal it, we want them to steal ours. They'll get sort of addicted, and then we'll somehow figure out how to collect sometime in the next decade."

-- Bill Gates during speech at University of Washington


"Number two market share goes to Windows pirated, or unlicensed ... That's a competitor that's tough to beat, they've got a good price and a heck of a product, but we're working on it."

-- Steve Ballmer, 2009 http://www.osnews.com/story/21035

Another quote by Ballmer, heard in person a couple of weeks ago: "In Russia, we have a great market share, but a bit of a problem collecting".

Sadly that didn't stop them from claiming that piracy was a $1.5B$ problem because the retail price of Office was $495. Even though even a child would tell you that no sane individual would spend more than 1/2 their annual salary to buy a copy of a buggy, but ubiquitous, set of tools. In fact they were willing to pay $2 for a pirated CD on the streets so the actual "loss to piracy" was $6M ($2 per each of those 3M machines) but that sounds so much less of an actual problem.

So, by this logic, BitTorrent piracy is a $0 problem? Doesn't sound right to me. Not that I'm really pushing the $495 figure either, mind you.

Actually yes, well actually approaching zero, and I realize it doesn't sound right. The cognitive dissonance comes from the economics of information, they work differently than the economics of goods.

Imagine this thought experiment. Imagine there was a 100% fool proof copy protection scheme in China, and the only way to acquire Microsoft Office would be to buy it. Try to imagine how many people, who make $1000 in a year would then cough up the $495 to buy it?

Its not hard to imagine because there is no manufactured good in China that commands a 'half annual salary' value proposition that is not critical to a person's survival. And Microsoft Office isn't that item.

There is an exemplar that there is a $6M value for Word based on the $2 CDs that sell, so its not 'valueless.' And there are benefits to owning a legitimate copy over a pirated copy. Further, the cost of producing a pirated product (box, CD, getting started guide, etc) is less than $2 so there actually a real-goods sort of economic value analysis possible as well.

So all we really know is that the number is >= $2 and < $495. And in information economics this poses an interesting challenge. How do you capture the maximum value for your information? Microsoft later introduced a plan where Office for this market was significantly less expensive than it was for the US and European markets [1]. That increased the value they extracted for office out of those markets tremendously. It was more than 10x that $1.50 price at $29/CD. And they captured a lot of value for it, millions and millions of dollars. But not billions of dollars.

Many have argued that much of NetFlix's streaming revenue has come from previous BitTorrent users [2]. Prior to NetFlix you could say BitTorrent was a "$xB" problem where $x is the money that has been re-captured by offering an alternative with an acceptable price.

But to understand how that works you have to think about what gives information value, its much more market driven that fiat driven. Sony can say "Its worth $X to own this movie" but if nobody buys it and it is heavily pirated they guessed wrong. Since the replication cost is nearly $0 they really are guessing at consumer value. Lots of DVDs sell for $9.99 or $4.99. At that price I won't even bother looking for a torrent, I'll just get the DVD. Its a market, it has different mechanisms that drive it.

[1] http://www.reuters.com/article/2008/09/23/microsoft-china-di...

[2] http://gigaom.com/2011/05/17/netflix-p2p-traffic/

  > Since the replication cost is nearly $0
  > they really are guessing at consumer value.
There still is the fixed cost of creating the DVD content. Maybe not encoding the movie, but putting together the menus, extras, easter eggs, packaging, etc. Also, the fixed costs grow when considering things like 'remastering vault copies.'

The variable costs might be close to $0, but the decision to create the DVD still costs something.

This is really interesting, thank you. Could you recommend any good books about the economics of information as you've described it here?

I don't know of any good books on the subject. I have been studying the ideas since 1995 when I joined a startup called GolfWeb which was publishing a magazine about Golf on the Web. And the question came up "What is this worth?" and "Would people subscribe?" That lead to some interesting investigations into the how people value things. Then in the late 90's I was talking with a friend who was getting an MBA at Berkeley. We invented the 'time valued coke machine' over lunch one day, its a coke machine where you can buy a soda for 25 cents if you are willing to wait an hour to have it come out. For $2 it will come out right away. Now this is a really really precise soda machine (has an atomic clock in it) so if you paid your 25 cents your ticket tells you exactly when a soda will fall out of it. Now you can sell that to someone for 50 cents, and if you're getting near the hour mark you might be able to sell it for $1 or maybe $1.50. So you have just created, and sold information for a profit of 100 - 500%. All when you thought you were buying a soda :-).

I started writing my experiences with this up but stopped when I went to work for Google on the advice of my lawyer. Its a discouraging thing to write/talk about because without context people think its crazy, so slowly over time, as context develops, and "real world" examples get played out, the notion that information can have its own set of economic principles becomes less foreign because it sure as heck isn't behaving like it "should" if it was bound by the old rules.

Anyway, three ways I've seen so far to add value to information is that its timely, its rare, or its distilled. An example of timely information would be stock trades, very valuable right at the time they happen to high frequency traders, less and less as it gets older until its basically worthless 15 minutes old. Continuing with the stock market 'rare' or 'precious' information might be insider information about a company, it has value because it creates context that others can't see, which gives it a value proportional to the advantage it conveys, and lastly distilled value. Remember our stock market trades? Well they were very valuable 20 milliseconds after they happened, worthless after 15 minutes, but when kept over the course of time new data can be distilled from that old information (like maybe a stock always rises in the third quarter). I discovered at a conference that the intelligence agencies call this last one the 'Mosaic effect' where any single piece of information is useless but together they allow new data to emerge.

Artistic creation however is always the "best" information in terms of value. It can create value greatly out of proportion to the cost to produce it.

Anyway, like I said, I've been thinking about this stuff for 15 years. :-)

It's a book that needs to be written; I hope you write it. (And then... how would you price it? ;) ) In the meantime, thanks for taking the time to dump this out--it's enormously appreciated and has given me lots to think about.

If people were selling copies on the street for $2, that suggests that $2 was roughly the right price point (in that area) to maximize profit, and Microsoft couldn't have made more money by selling it for more.

That argument ceases to hold in the case of BitTorrent, where the distributors aren't trying to maximize their profit (or make any at all).

See previous response, Microsoft actually did change their pricing and captured a huge chunk of value. It was significantly higher than $2/CD. When selling information there is a tail, a few people will pay whatever you ask, a few more will pay less than that, Etc, until you get to all of the people who would obtain a copy if it were zero cost. That defines the "value curve" of the information. It is not a straight line. Because it is "hard" to justify large differences in price for the same product, one of two strategies are used a "steadily decreasing price" or a "fixed but relatively low price". Depending on the product different things make more sense. Games for example benefit from the 'expensive at first but cheaper later' model, reference works benefit by capturing as much market as possible quickly.

It's similar to many startups. Free initially to ensure growth, but always with eventual monetization in mind.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact