Fog Creek Software
Discussion Board




Economics of Porting Software to other OSes

In http://www.joelonsoftware.com/articles/fog0000000051.html Joel argues that the costs of porting a Software to OS X have to be less than 10% of the original development costs because otherwise it would not make sense because that 10% is Apple's market share.

I'd like to point out that this is flawed logic, because he is making connections that should not be made. For a business decision like porting software, there are really only two factors: investment and return. So if original development costs were 5 million, and costs of porting would be 1 Million (20%) and you could sell the software to 1 Million Mac Users at 2 $ each, it would be a bad decision not to do it, even though the costs of porting are 20%.

This is a mistake often made. Consider the following: If you just bought a car for 20000$ and a book for 50$ and around the corner you find a show which offers the same car for 19960$ and the book for 30$ and you could return both the book and the car for a full refund, which one are you more likely to  return? (Normally, you do this test with two groups. Usually, most would return the book but none would return the car). But in both cases you could have saved 20$ for the same amount of work.

To get back to Joel's text: It is also wrong to assume that you would get the same marketshare on OS X as on Windows. Since there is a lot less Software offered and Mac Users tend to have more money and work professionally, there is a real possibility that the market share would be significantly higher.

Just some thought, enjoy your sunday.

Matthias Winkelmann
Sunday, June 06, 2004

Yes there is an assumption to Joel's logic that the same percentage of Mac users would buy the software as the Windows version. This may be slighty flawed, but as a quick "on the back of a napkin" calculation it's probably okay.

You do know that Joel has actually ported Fogbugz to Linux and OS/X don't you. I wonder if the same percentage of users did buy Fogbugz on Mac and Linux as the original Windows port?

Matthew Lock
Sunday, June 06, 2004

I believe Joel's estimate of a 10% Mac OS market share already includes assumptions like "Mac users have more money" and "Mac users buy more software". The 10% are not the real market share but the value to work with in his calculation. Call it "effective market share" if you will.

Leauki (Andrew J. Brehm)
Sunday, June 06, 2004

Just consider the difference between the ratio of piracy in the Windows and Mac software PLUS the difference between the average purchase power of the Windows and Mac users.

Porting software to target Linux wouldn't be as rewarding. But for Mac and Unix (the paid one) market. Yes.

Green Pajamas
Sunday, June 06, 2004

Ofcourse it depends the kind of software. If Linux markets pays for it. Sure.

Green Pajamas
Sunday, June 06, 2004

The original poster ignores the opportunity cost of porting to OS X.  If Joel spends time and money writing a Mac version, he's not spending that time and money improving the Windows version or making new software.  It's only a "bad idea" not to do the OS X port if (a) the OS X port will make money __AND__ (b) the OS X port will make MORE money than a comparable investment of time and money in another area.

Emperor Norton
Sunday, June 06, 2004

To be fair, Joel's article wasn't really about the profitability of porting software.  It was included as an aside to point out why REALBasic should be more VB-compliant (in order to reduce the cost of porting software to the Mac).

In that context, it's not unreasonable to make an oversimplified assessment of the situation.  Clearly if you're making a business decision like that, you'll have to weigh all the factors, including market share, opportunity cost, and strategy.  There's no quick 'n easy formula...

Joe
Sunday, June 06, 2004

*  Recent Topics

*  Fog Creek Home