converting capital into what???

greenspun.com : LUSENET : Joel on Software : One Thread

"Imagine that the goal of your software company is not to solve some specific problem, but to be able to convert money to code through programmers."

Yikes. Software exists to solve problems. Software companies that don't solve problems fail. Code that works is either the consquence of business plans that work or of charitable creative types (like opens source developers). Many technical markets are actually content or service markets. AOL is content. Dell is service. Etc.

Three false observations:

1. "Back when the browser integrated with Windows was crap, Netscape had 80% market share. So please stop fretting about the power of bundling."

Yes, stop fretting about it and start using it (if you can). *Had* is the key word here. Bundling was THE reason that MS trumped Netscape. Netscape lost out on the Christmas upgrade/release cycle due to MS games, and MS won. End of story.

2. "It's amazing how many companies - Lotus, Netscape, WordPerfect - have bitten the dust because they lost the ability to convert capital into code."

Each of them were trumped by Microsoft b/c they lacked the ability to bundle with Windows. As the author noted, Excel is mostly used for keeping lists. Word is mostly used for simple notes that could use Wordpad. But you need word b/c someone else has a word binary. You get IE b/c it works best with MS. You get Excel b/c it comes with Word. Bottom line -- bundle to breed your market.

3. "Hire smart people, and they will produce good stuff that you can sell and make money off. Then everything else follows."

Right. Build it and they will come. Only if you tapped a market need.

Conclusion: Business people should build businesses. Technical people should build software to serve them. Bill and Steve and Larry and all the rest cared not about their code. They built their empires on market *opportunity*. Seek and ye shall find. And let the rest erect their monuments in the desert . . .

-- Anonymous, March 11, 2001

Answers

3. "Hire smart people, and they will produce good stuff that you can sell and make money off. Then everything else follows."
Right. Build it and they will come. Only if you tapped a market need.

Perhaps you should take a look at Autodesk's history, as written by one of their founders, before you go dissing this idea, or proclaiming:

Business people should build businesses. Technical people should build software to serve them.

Or, for that matter, read up on HP's history. Both companies were driven by technical people who wanted to solve problems, and found a market among people ith problems. AutoDesk in particular is an interesting case study, since it's a pure exercise in code Darwinism - chuck a bunch of technical people to write what they like and see what sticks.

(You'll also note that various "business people" lost out on a chance to get a slice of the AutoDesk action).



-- Anonymous, March 11, 2001

AutoDesk is a compelling story, I agree. But the Darwinian approach is expensive -- how many kangaroops came and went before the world got a kangaroo? I think writing code willy-nilly is great for the world -- I used to do it when I had the resources -- but if you want to be serious about money identify a market need *first* and then build a software solution to serve it profitably.

Note: Technical ppl write software to serve the *business* NOT the business *people*!!! (The business people also serve the business, but directly rather than via their code. In the end, we all work for the invisible hand).

-- Anonymous, March 12, 2001


I don't think anyone is advocating the development of software for software's sake - the open source guys have that sewn up!

I think it is disengenuous of business to pretend that they can accurately identify and quantify market opportunities then followup with an attack strategy that guarantees 10:1 or 100:1 return to investors. This is the current climate - an I admitt to have written a few of these business plans

IMO, it is very bourgiose and protestant to undertake a venture in this manner. Amass captial, build, sell, cashout. Too simple and not enough homage paid to the steps inbetween. That's what I read in Joel's article. Reminds me of the underwear gnomes from southpark: Step 1. - collect all the underpants Step 2. - Nobody knows Step 3. **profit**

It business seems to want to treat software development as a production facility that can be assembled from interchangable nerds (which they can treat like nerds). And since the factors that influence productivity are lost on "non-participants" it is easy to see whay that happens.

Creating a great software development engine is hard to do, takes time and is very human. Those things don't fit well into business plan.

-- Anonymous, April 24, 2001


Moderation questions? read the FAQ