Brad Feld

Month: November 2005

In my last post in the Business Plan series, you discovered that my first company was a “software company.”  What kind?  Even in 1987, the software industry covered a wide variety of things.  As I mentioned before, the Introduction section of a business plan should start with a general overview (“microcomputer software”) and then get specific (“database stuff”).  By the end of this section, the market segment you are targeting (in this case – “semi-custom software”) should be defined.

The early days of microcomputing were plagued by a lack of software. The hardware component of the system evolved the quickest; software developers were always playing catch-up. Support tools for software companies were limited forcing the programmer to invest a significant amount of time in developing toolboxes for use on a particular machine. As a result, software developers often sacrifice the flexibility of the applications they were developing in order to simply get them to work. This changed with the advent of the IBM PC which prompted an epidemic of fourth-generation application development languages, more commonly known as database languages. For the first time, the microcomputer software developer could focus on the issues underlying his application instead of spending all his time trying to implement the application on a particular piece of hardware.

Today, database languages are beginning to replace conventional languages within the context of application development. The trend has reached a critical mass; new database languages are emerging weekly. This outbreak of products has served to legitimize a new approach to application development.

These database programming languages can serve as a basis for a new type of application development – semi-custom software. Semi-custom software is a divergence from existing software product offerings. In a semi-custom environment, the benefits of mass-produced systems and custom software are combined. A reusable “shell” is designed for a specific industry – this is the systems component. Instead of packaging and selling only the shell, a semi-custom company modifies it to fit the client’s specific needs. Some custom software is written and integrated with mass-produced software. As a result, the customer essentially gets a software product that fits his precise needs (emulating a custom product) at a systems software price.

In 1987, “semi-customer software” was a new concept.  Database languages (or 4GLs) were becoming popular (remember dBase and .dbf files?) and were starting to incorporate mainstream programming capabilities (most notably procedural abstraction – a big deal at the time).  The things we now call “packaged applications” were going by pre-ERP TLAs such as MRP, CAD, CAM, and POS (“point-of-sale”, although most were about as good as the other use of the acronym).  We felt like there must be something in between a custom application and shrink wrapped software and decided to try to coin the phrase “semi-custom software.”  While this phrase didn’t stick, the concept ended up being very relevant and foreshadowed the packaged software revolution.


Kimbal Musk just put up a blog about his brother Elon’s rocket launch that is scheduled for tomorrow on Kwajalein Atoll.  I wrote about it yesterday – Kimbal has a lot more pictures up and I expect he’ll be reporting live (or hopefully simply “alive”) from the scene tomorrow.


Jim Collins has an outstanding short essay titled “Lessons From A Student Of Life” as a tribute to Peter Drucker in the 11/28/05 Business Week.


Stephen Wolfram has accomplished some remarkable things in his life, including creating Mathematica, a very successful private company called Wolfram Research, a set of amazing mathematics web sites including The Wolfram Integrator, and an overwhelming tome called A New Kind of Science which I intended to read last summer but instead just stared at it each day.

Now Wolfram brings us Ringtones (well – WolframTones).  If you ever wondered about the math behind ringtones, now is your chance to play around and create your own ringtones using “simple programs from Wolfram’s computational universe, music theory, and Mathemetica algorithms.” 

Nerd heaven.


In my last post in the Business Plan series, I promised to tell you about “The Industry”.  This is an important early section of a business plan that frames the overall industry the company is part of.  It’s important to keep this section short (if I want to learn the history of an industry, I’ll read a history book, not a business plan.)  The first part (or first few paragraphs) should describe the industry generally and then become more specific about the segment of the industry that the company is going to address.

In 1987, people talked about “the microcomputer software industry.”  The gorillas of this business were Microsoft and Lotus – gigantic software companies weighing in at $200 million in annual revenue.  There were many smaller companies (Ashton-Tate, Aldus, T/Maker, MicroPro, WordPerfect anyone) – of which today’s generation of entrepreneurs has never heard.  Following is how I described the microcomputer software industry in my business plan in 1987.

In the ten years since Microsoft introduced the world to personal computer software via Microsoft Basic, the microcomputer software industry has grown from nothing to a multi-billion dollar business. In the wake of this growth are thousands of software companies ranging from garage operations to $200 million giants such as Microsoft and Lotus. For a while, thousands of people became rich overnight using a simple formula – create a new piece of software and toss it out into the market via magazine ads and user groups. Software heroes were common – all one needed was something neat. Even the high school computer hacker got in on the action (much to the amazement of his parents who soon were making less money working full time) by spending his afternoons writing a game and selling it to an established company.

It was inevitable that an industry (this industry is narrowly defined as the microcomputer software industry – specifically companies writing software for IBM PC, PC compatible, and Apple computers) growing this rapidly would attract some hungry, experienced capitalists. These people took the form of senior engineers, venture capitalists, and MBAs. As the pool became more populated, the structure became increasingly chaotic. No longer was simply anyone able to succeed – competition began to play a significant factor. The rest of the business world took notice as software companies began going public, large companies started divisions that developed software, and Business Week ran feature articles on the software revolution.

As 1987 begins, the microcomputer software industry is entering adolescence. The organizational frenzy of the past is becoming less of a factor. Do not interpret this as a slowdown in activity – the software industry is busier than ever. It has merely taken on some structure. Methodologies for success are being established and are becoming the norm. The days of easy money for everyone are over.

If you substitute “Internet” or “Web 2.0” for “microcomputer software industry”, does is still work?  (Entertainingly, whenever I try to type “microcomputer software” I end up typing microsoft first.)


The War Room

Nov 23, 2005

Think back to 1992.  “It’s the economy, stupid!”  Amy and I watched The War Room tonight, the brilliant documentary of Bill Clinton’s 1992 presidential campaign and the organization behind it, spearheaded by James Carville and George Stephanopoulos.  Carville’s quote was actually “the economy, stupid” and was part of a haiku that Carville came up with.

Change vs. more of the same
The economy, stupid
Don’t forget health care.

Carville is an absolute genius and unbelievably entertaining as a special bonus feature.  I’d team up with that dude any day just for the laughs.  Stephanopoulos reminded me of Sam Seaborn which – it turns out – was intentional.


If you haven’t used Technorati for a while because of either performance or accuracy issues, try it again.  Dave Sifry – Technorati’s CEO – has a post up describing their performance improvements.  As a Technorati investor, I’ve given Dave and team lots of steady feedback and have watched with happiness as the performance, relevance, and accuracy has steadily improved as they’ve continued to scale up and tune their infrastructure to handle the massive number of blogs that they are indexing.  Oh – and they’ve added a bunch of cool new features along the way.  Dave and his team listen – if you have issues after trying it again, email me and I’ll pass it on or give them feedback directly.


If you want something chewy to read over Thanksgiving, I recommend FeedBurner’s thoughtful piece titled How feeds will change the way content is distributed, valued and consumed (also available as a PDF.)  This is the first article in FeedBurner’s “Feed For Thought” market reports.  Fred Wilson also had a good post on The Second Coming of RSS worth chewing on.  I promise neither of them are turkeys.


On Friday 11/25, Elon Musk’s new company, SpaceX will launch their first rocket, the Falcon 1, at 1pm PST.  I got a note from Elon’s brother Kimball – who now runs an awesome restaurant in Boulder called The Kitchen – with some photos.

Yes – that’s a rocket ship on a remote island called Kwajalein Atoll.  According to Kimball, Kwaj is the largest Atoll in the world, 1,400 miles away from Guam, 2,100 miles away from Hawaii, has a population of 2,500, a runway, a small military base, and a lot of excited rocket scientists hanging around.