Slashdot Mirror
Ten Geek Business Myths
hpcanswers writes "Venture capitalist Ron Garret has posted a list of eleven (despite the title) common mistakes entrepreneurs with a technology background make. A common theme is that good ideas sell; in reality, what a customer wants sells. By extension, having a Ph.D. and holding a patent are not particularly helpful if the intended end-user does not have the same level of understanding of the widget as the creator does."
In case of a /.ing
Myth #1: A brilliant idea will make you rich.
Myth #2: If you build it they will come.
Myth #3: Someone will steal your idea if you don't protect it.
Myth #4: What you think matters.
Myth #5: Financial models are bogus.
Myth #6: What you know matters more than who you know.
Myth #7: A Ph.D. means something.
Myth #7: I need $5 million to start my business
Myth #8: The idea is the most important part of my business plan.
Myth #9: Having no competition is a good thing.
Myth #10: After the IPO I'll be happy.
It's ironic. The two things that make engineers so good at engineering are the two things that make them so unsuited to running a business.
Hubris is a trait of engineers that makes them strive for greatness in their products. After all, you can't really have good pride if you're constantly getting negative reactions to your stuff. However, it also leads to a close-mindedness and tunnel-vision in regards to other technologies and solutions. A good businessman must be able to survey the market and understand the positioning of his product. Someone who thinks that they have such a great solution that it is applicable to any and all problem domains is selling snake oil. See Netscape and Sun's Java for two examples of solutions that were billed as much more than they realistically were.
Laziness is a good trait for engineers because it forces them to seek efficient, easily-implementable solutions to everyday problems. Automating tasks is absolutely essential to creating value in a company. However, the business side of running a business is not reduceable to a script. There are serious tradeoffs that must be weighed all the time in order to guide a business down the road to success. These can't be automated. The laziness trait leads engineers to seek easy solutions when they should be seeking difficult-to-find synergies. Well-designed software is modular with simple interfaces. Well-run businesses are well-integrated and derive their strength from business units coordinating with each other, not simply acting as a pipeline from one end to another.
Perhaps you would like to step into the role of a PhD who *does* perform research and development? I've found that most of the time when people say stuff like this, they have no real idea of what is involved in either obtaining a PhD or working as one.
The PhD not only demonstrates that you are capable of thinking critically, it shows that one is able to communicate, analyze and create new "content" and make advancements. Speaking as a PhD, the job is much harder than I ever thought, though it is fun and I would not do anything differently. Having to write grants, write papers, teach, perform science, deal with administrative duties all at the same time is a much harder job than most folks realize. Of course that is just academia. If you add in work in the private sector on top of that, you have even more responsibilities (though prospects for more money). Some PhDs of course stick to industry and do quite well. That's all fine and dandy, I just like the additional challenge of academics in addition to commercial work.
Visit Jonesblog and say hello.
- Sales team. The first person you'll ever need is the VP for sales. Then you'll need a good, experienced sales team. No matter how expensive they are, these people will make or break a business. I ended up leaving my job and joining a startup precisely for this reason.
- Time frame and financial needs. One thing all startups underestimate is the need for quality assurance. Generally, testing for defects takes more time than assembling a product. Thus, the time to market should be at least tripled and the cost doubled from what you expect.
- Intellectual property. True, patent protection is overrated. However, there are thousands of inventors and companies waiting to sue your ass off if you infringe on their patents. More important than filing your patent is to research whether you infringe on others' patents or not, and settle any licensing issues. This will get very costly, and in this case getting good lawyers is worth their weight in gold. We pay roughly $5000 per patent examined, but they decreased the number of patents we thought we would license from 40 to 2.
- Company share. Many inventors don't want to relinquish control over the company, and want to maintain a majority stake at any cost. Most investors wouldn't agree with that, with a good reason - a researcher running a company is recipe for disaster. And as the classic saying goes, it's better to have 5% of $100 million than 100% of nothing.
- IP ownership. I talked about infringing on others' IP already, but what about the inventor's IP? The inventor must transfer all the rights to the invention to the company. Otherwise, the inventor will exercise undue influence over the business, and sooner or later (rather sooner) this will create conflicts between the inventor and the management.
There are plenty more rules of the game, but this game is too flexible to make any of them universal. The best thing is to give over your technology to a seasoned entrepreneur and just ride along.Why should I believe you? Do you have a Ph.D?
- RG>
Hey pal, this isn't a pleasantforest, so don't waste my time with pleasantries!
Well, in the sense they were available, yes, but the guy was talking about Microsoft going one step further and bundling it. I never owned an Atari ST, but I can tell you that none of the TCP/IP stacks for the Amiga were ever bundled with it in Commodore's lifetime.
AS225, Commodore's own stack, was never even released outside of a handful of developers. Amiga users had to rely upon AmiTCP or KA9Q (renamed "AmigaNOS") to get it to work. AmiTCP was free software (BSD license, IIRC) up until the 3.0 betas, but controvertially went shareware with version 4. KA9Q was... uh... yeah. You didn't want to use it.
To go on to other mainstream platforms of the time: So far as I'm aware, it wasn't until the late nineties that MacOS had a stack bundled with it. Stacks were available before 1990, largely due to the Mac's entrenchment in academia, but they weren't bundled with the system. OS/2 Warp 3 "came with" a TCP/IP "stack", but for consumer versions it was close to useless. It only supported SLIP, and wasn't modular, so you couldn't just add a device driver for your Ethernet card (or just PPP) and it'd work, you'd have to throw the entire stack out and buy the premium version from IBM.
So really, other than Unix, no mainstream operating systems came bundled with a full TCP/IP stack until Windows 95 did. I hate to say it, and maybe it'd have happened anyway, but Microsoft did pioneer there.
You are not alone. This is not normal. None of this is normal.
I have to agree. Vulture capitalists can be the worst offenders when it comes to stealing ideas. They will never agree to any kind of NDA, and if they like what you present, but not your team, they'll sometimes go build a team on their own to pursue your idea. Of course, if they really do hate your team, you probably will fail anyway.
Those were all really good points in the article. I've personally stumbled into most of those traps (being the prototypical geek). Just for fun, I'll list some of my mistakes that correspond to his points:
#1 - A brilliant idea will make you rich.
In 1991, I started DataDraw as a company on the side that would sell software to make teams of programmers more productive. It's a great idea, with huge potential to benefit the whole planet. All it requires is that all those programmers out there understand how they can be more productive, care, and then take action to change. The harsh reality: they don't figure it out (go read datadraw.sf.net if you think you're really smart); they don't care (it's all just money after all); they don't like to change (show me your computer language of choice, and I'll guess your age within 5 years).
#2 - If you build it they will come.
Err... see #1. The next company I started also suffered from this problem. A friend and I started OpenASIC to solve the terrible communication problems between EDA tools. I wrote a very complex and fairly complete LPM module generator, simulator, and various readers and writers. My problem this time was that I BELIEVED what the customers were saying. Just because every major EDA and FPGA company issued press releases supporting LPM doesn't mean that they actually want anything to do with it (it's basically now an Altera specific format). Learning the difference between what a customer will buy, and what he says he will buy is key.
#3 - Someone will steal your idea if you don't protect it
I've had ideas stolen by professors and managers, and I've been stiffed by clients who decided not to pay me after they learned all they needed. The underhanded BS that happens in startups is unreal. Stop worrying about protecting your ideas, and worry about the guy who's gonna try and steal you blind.
#4 - What you think matters
I agree and disagree with this point. Many geeks imagine that if they like something, then so will customers. That's just plain wrong. However, if you actually listen to the customers, and go build what they ask for, you're sure to go broke. You have to be like Steve Jobs, and figure out that people want to pay more for a music player (not less), and that looking cool, and being bone-head easy (so dummies can use it) is what counts. You wont get average customers describing themselves as vain and stupid, but you'd better understand that most people are!
I started a company in 1996 called FPGA Technologies, with the purpose of creating embedded FPGA IP for SoC applications. I listened to all the SoC guys complaining about rising tooling costs, and heard their very enthusiastic response to my proposed FPGA cores. So, I went and built it... and got out when I realized that the customers were wrong. FPGA cores are waaaay to big to make sense in SoC applications. Stupid vulture capitalists keep on funding these poor doomed startups that want to do the FPGA IP thing. It makes a great elevator pitch, but a lousy product. The latest is M2000, which will most likely go broke when investors get fed up with them.
#8 - I need $5 million to start my business
That's funny coming from a VC, since few VCs will consider investing in a company that needs less than $10M to go public. They often have hundreds of millions of dollars to invest, and they can't waste time tracking every $1M investment.
However, I believe there's a huge opportunity for geeks like us to get semi-rich doing non-VC funded startups. In 2000, I moved to North Carolina, and started ViASIC. We have some angel investors, but no VCs, and our investment to date has been qu
Beer is proof that God loves us, and wants us to be happy.