Slashdot Mirror
Make More Mistakes
prostoalex writes "Eric Sink from SourceGear, well-known in the open-source world for AbiWord Suite, shares his thoughts on starting and running a software business. One of the keys to a successful business, as Thomas J. Watson once said, was to double the rate of failures. Eric Sink tells the story of what mistakes he personally made, what could be avoided, and what's important for geeks to know when starting a software company."
Companies like Intel, HP etc can make blunder after blunder but still come out on top as they have the resources to "wait" for the winner, and more importantly put a lot of resources behind the winner once invented.
Take Intel's Itanic, or the 860 this would have sunk any company but the very large. Intel's Yamhill is waiting in the wing in case The Custoenmrs want it How may companies can do that.
The Venture Capital stragegy centers around this Throw Mud on the wall and see what sticks and should more rightly be seen as Outsourced R&D that business start-up. FYI, More Start-ups get absorbed pre IPO than go public.
Help fight continental drift.
...and now doing as much service work as development, I have success and failure on a daily basis. It has been both frustrating and exhilarating, but I learn at a rate much faster than when I was in development alone. Development is paced with a chance at total failure at each release. In my service role, I fly by the seat of my pants and can fail on a daily basis (and succeed). I would compare it to the speed of insect evolution compared to mammals.
Below listed are some of the features hidden by the writer, perhaps guided by principles of neurolinguistic programming or freudian psychology... err.
Be careful about using bleeding-edge technologies.
i.e. use established (i.e. Microsoft) technologies.
A market with no competition ain't.
i.e. stay away from free software. (as microsoft always propose that free software is anti-competitive)
Small ISVs should build apps, not platforms.
i.e. build with existing platforms (i.e. microsoftish)
You could also try the OpenSource method of making a software business work, which is to release all of your mistakes and failures to the public with full annotation and then let them write the next version of the software while you sell 'support services' for it to the general populace.
Works a treat for service products like Apache, where the service the software provides is the actual value of it, and the act of releasing the code to those interested simply means more patchers working on the problems. Likewise it works for common software like Office software, where the people who use it don't generally care to understand the code.
Ironically, it fails in products targetted for sale to code-competent people, as this is a market that can support their own needs, but then this is a small market who could code their own product rather than pay any money anyway.
An infinite number of monkeys will eventually come up with the complete works of
Its quite evident that this guy had lots of money. Lots of money makes it easy to make the bone-headed mistakes this guy made. He bought and office building!!!!! This guy does not belong at the top of a company. Don't listen to him. I created a company Rockstar Software with $8000 seven years ago. I know from where I speak. This guy is a total moron. Run away from his advice. Run far far far away.
Increasing your rate of decisions made may increase the number of failures. But it should also increase the number of successes.
There is no point in sleeping on a decision when their will be another bucket load of decisions to be made tomorrow.
One issue I had with the article was this comment:
"At the risk of being too obvious, let us observe that every ISV is started by an entrepreneur who somehow overcomes fear of failure."
I think the fear of paperwork is probably just as powerful a deterrent as fear itself. I know that I'd rather have a software project bomb than have to deal with all the forms and paperwork running a business generally entails.
I think having many failures is a symptom of the important trait of persistence more than a prerequisite for success.
Let's take someone who has done well with a particular business. They can become complacent, rot and die. Or, they can pick out what was behind the success, suss out what was plain dumb luck or the result of excellent ideas, planning and execution. Then follow up by doing more of those good things.
Yes, I know that the article was published on MSDN, so all of the open source zealots can start in with the conspiracy theories about Microsoft's involvement with publishing it. But the fact remains that an experienced proponent of open source no longer views it as a viable business model and has abandoned his efforts to make money developing, selling, and supporting open source software. You can make money selling hardware that incorporates open source software/firmware. You can sometimes make money supporting open source software. But it's damned near impossible to make money by developing an open source product and selling/supporting it.
Now the nuts can come out of the virtual woodwork and start screeching about the one true religion of open source, but he fact is that not one in a hundred thousand of them has successfully started a major corporation that develops and sells open source software. So if you want to claim that all of those venture capitalists, along with Eric Sink, the developer of Abiword and the founder of SourceGear, are wrong, please include your business credentials when you reply to this.
>Below listed are some of the features hidden by >the writer, perhaps guided by principles of >neurolinguistic programming or freudian >psychology... err.
>Be careful about using bleeding-edge >technologies.
>i.e. use established (i.e. Microsoft) >technologies.
No. Microsoft is always releasing new technologies, until they are commonly accepted/rejected they are bleeding edge.
>A market with no competition ain't.
>i.e. stay away from free software. (as >microsoft always propose that free software is >anti-competitive)
No. He simply states that if someone else isn't doing it, there probably isn't a market for it.
>Small ISVs should build apps, not platforms.
>i.e. build with existing platforms (i.e. microsoftish)
Why not *nux or mac based platforms. He is more or less saying if your building a small ISV, don't re-invent the wheel.
Nothing like the mention of Microsoft in an article to bring out some good Slashdot group think.
this article is a good read for anyone thinking or in the early stages - we've been there, done that especially:
1) betting on java (bleeding edge) for a gui based program
2) spending too much on office space
Eric is right on about reading - key on the books where people tell their war stories and software books like "Microsoft Secrets", "Inside Intuit", etc - my favorite book is Charlie Ferguson's book "High Stakes No Prisoners" especially where he talks about spending 3 months of decision tree analysis - looking back, that could have saved me $300,000 that I couldn't afford to lose - always have backup plans for all the mistakes you are going to make, or when external things go against you.
And btw - notice most of these stories have pretty smart Ivy League and MIT guys telling you how hard it is and how much they screwed up. In short, it's very hard, odds are against you, you must be very, very careful, and you must keep money coming in no matter what you have to do because you must use the 3x rule: plan carefully how long or how much it will take and multiply by 3 - uh huh, works out about right. And yes, don't drink the coffee, sell it - it's much more profitable.
"Robert Scoble, weblogger extraordinaire, recently said, "I want to see more software companies, not fewer." I heartily agree."
I'm guessing Bill Gates isn't the author here.
Pelé!
If your'e starting a business, maybe working your own ass off will help. If you're working in one, probably not.
By the last set of figures I saw, the US currently ranks #1 in number of hours worked in the industrialized world, but only #3 in per person productivity. (The average US worker produces a contribution to the GNP of about 30 $ US per hour, while Great Britain was 50 cents (again US) or so higher, and the reunified Germany had assimilated the east German economy and was back up to about 33 $ US per worker hour.
These are estimates circa 2002, and no I don't remember just where I read them, probably some scandal sheet like the Wall St. Journal.
The interesting thing was Germany had just made it manditory for most businesses to give DEC 26th off to facilitate people traveling to be with familys over the Christmas holiday. Many business owners had complained that the German economy would suffer from this and other new increases in non-work time (new maternity leave rules, increased minimums for vacation weeks/years employed, and such). Instead, the economy had apparently benefited.
So if you want to be a successful creator of a new company, by all means work your tail off, put in those 16 hour days and six week stretches, but hire people who want to work smarter not harder.
Who is John Cabal?
Clearly you are speaking of yesterday's market. The software consultancy I work for (which charges at least hundreds of thousands of dollars) is just finishing our best year ever.
Most companies slashed IT spending in 2001-2003, certainly. Most software companies that appeared in the boom died (thankfully, as most of them shouldn't have existed in the first place). 2004 shows every sign of loosening the purse strings, and upgrading all that software that's been getting stale.
As to globalization, we're experiencing the same thing that the US did at the turn of the last century, when 30% of Americans were farmers. Now it's 3%. Guess what? We survived, even though it wasn't a pleasant re-tooling. Sure software jobs are moving overseas, as are manufacturing jobs (even China LOST 15 million manufacturing jobs in the last ten years, to greater efficiency in the plants). If you can't adapt, get the hell out of the market and open a gym, a bar, a restaurant, even a stripe (sic) club.
Certainly, if you provide a mediocre software product or service, 2003 wasn't a good year for you. If you were at the top of the software profession, it was just fine, thank you very much. Don't blame the economy for your own failures.
A difficult year marked by slight hope
Record losses at German blue chips, but restructuring and rationalization begin to show their effect
january
Economics Minister Wolfgang Clement issues a special ministerial permit to allow the takeover of Ruhrgas, Germany's main natural gas supplier, by Eon, one of Germany's two dominant electric utilities. The ministerial intervention overrules the Federal Cartel Office, which had warned against impaired competition both in the electricity and gas markets.
february
The stock market collapse, record insolvencies and belated restructuring and rationalization efforts have plunged German banks into a crisis. Commerzbank and Hypo-Vereinsbank both post the first annual losses in their corporate history. Experts predict drastic sectoral consolidation.
march
At 4.7 million, unemployment reaches the third highest level since unification. The jobless rate stands at 11.3 percent.
Dresdner Bank Chairman Bernd Fahrholz is sent packing as parent company Allianz publishes the first annual loss in its corporate history for 2002, with Dresdner being the biggest burden.
Deutsche Telekom posts a record loss of EUR24.6 billion for fiscal 2002, the highest loss ever posted by a German company.
Wella's founding family agrees to sell the world's second-largest maker of hair-grooming products to Procter & Gamble.
The Bundestag decides to extend shop opening hours to 8 p.m. on Saturdays. The new regulation will take force on June 1.
april
In their spring forecast, Germany's leading economic institutes project economic growth of 0.5 percent for 2003, revising downward their earlier forecast of 1.4 percent. The six think tanks expect the German deficit to reach 3.4 percent, exceeding the limit of the euro-zone Stability and Growth Pact. The government remains optimistic and issues only a slight downward revision of its growth forecast to 0.75 percent from 1 percent.
Germany's most powerful industrial union, IG Metall, reshuffles its leadership. In a surprise move, the board nominates deputy head Jurgen Peters, a hardliner and ardent defender of Germany's extensive system of worker protection, as the successor to Klaus Zwickel.
Frankfurt airport operator Fraport cancels its dividend and discloses a net loss of EUR120 million for 2002 after writing off an ill-starred airport project in Manila launched in partnership with business cronies of the discredited former ruler of the Philippines.
may
The German economy slipped into recession in the first quarter of 2003. Finance Minister Hans Eichel publicly abandons his longtime goal of balancing the federal budget by 2006.
The level of management pay in Germany becomes a subject of public debate. Federal Justice Minister Brigitte Zypries of the Social Democratic Party demands the disclosure of individual board member salaries to improve investor protection.
WestLB posts a record loss for its 2002 business year after admitting that it had to increase risk provisions for its London project-financing arm over a risky deal with British television and radio leasing company Boxclever.
june
Deutsche Borse closes the badly tainted Neuer Markt segment for young and supposedly fast-growing companies. A new, untarnished Tecdax index now serves as the benchmark for investors in stocks that would have been called new economy a few years ago.
The collapse of life insurer Mannheimer Lebensversicherung becomes a first test of sectoral rescue company Protektor, which takes over all 345,000 contracts.
West LB's multi-billion loss causes heads to roll. Public prosecutors investigate both its London group and several managers. Chairman Jurgen Sengera steps down, making way for interim Chairman Johannes Ringel.
Robert Bosch acquires a majority of heating equipment maker Buderus, making Bosch the European market leader in this segment.
Quelle becomes the first German mail-order company to sell cars over the Internet.
After four weeks of industrial action, IG Metall boss Kl
And even more ironically, after poo-pooing "bleeding edge" technologies, he's recommending .Net. Pot, meet kettle.