More silly anecdotal hand waving from Techcrunch guest writer Glenn Kelman and Guy Kawasaki about serial entrepreneurs.  They seem to argue that it’s all luck and that serial entrepreneurs suffer from so much hubris that they can’t help but do worse the second time around.  The argument goes that Amazon, Apple, Dell, Ebay, Google, Microsoft, Oracle and Yahoo! were all founded by first timers under the age of 30.  If these guys are right, the VC’s should fire most of their partners and buy Ouija Boards because if past performance is no indicator of future results, it’s all luck for them too.  Limited Partners need to rely on random selection and place a lot less money in the hands of a lot more funds to gain a better portfolio effect.

But that doesn’t seem to be the case.  Returns for venture funds are correlated with prior results.  Likewise, serial entrepreneur success rates are better than first timers according to one Harvard Study.  There are certainly examples of serial entreprenuers that have done more than one successful thing.  Marc Andreesen is one such that I’ll throw out.  My old boss Philippe Kahn has had 3 large successes so far, 2 of them post Borland, and is hard at work on his fourth.  There are others. 

That the list of largest, most successful tech companies like Microsoft and Google is largely about first timers should come as no surprise.  Hitting the Uber Home Run is much more a matter of luck than skill.  You can’t set out planning to have the next Google sized success.  One of the Valley’s leading VC’s lamented to me recently that the VC business is turning in disappointing returns precisely because it got used to the idea that Google’s could be made, and it turned out they come along once a decade or worse.  Ignore those cases.  They are not reality.  They are are the exceptions.  Companies don’t have to create Google-sized results to be interesting.

There is something else to consider.  Luck was an essential ingredient in the Microsoft, Google, et al story, but there was another essential ingredient too.  To argue that these companies were all luck is a ridiculous insult to people like Bill Gates, Jeff Bezos, Michael Dell, Larry Ellison, and Steve Jobs.  It took enormous luck combined with enormous talent to bring about these lasting successes.  Consider Steve Jobs.  Do we really think he hasn’t proved himself more than once and that his career is all about luck and youthful enthusiasm?

Glenn Kelman in his Techcrunch article says a key thing, but his conclusion about what it means is all wrong:

But to be really great, I wonder if second-timers have to forget some of what it cost us so much to learn.

The conclusion is that once they learn what works, the second-timers will be prisoners to that formula. 

The mistake is in not realizing that real entrepreneurs never stop learning.  They don’t assume that there is one formula that can be applied over and over with success.  There isn’t.  You have to step out of formulas, get above it all, and figure out how to get inside the decision loop of the markets.  What does that mean?  It’s a topic worthy of a blog post unto itself, but to understand you need to learn the OODA theories pioneered by jet fighter pilot John Boyd.  Amazingly, they are exactly on target as effective business strategy.

What serially successful entrepreneurs do is understand the big trends much more quickly than the market at large.  They get there faster than the rest of the market and hang on to the valuable leadership they’ve staked out tenaciously.  This is not an endorsement of the first mover approach.  It is a recognition that business competition has a lot in common with an aerial dogfight:  the other guys have talent (and often more resources) and the stakes are frighteningly high.  If you can “get inside” your opponent’s decision cycle, you are in the driver’s seat.  They are responding to your initiatives and this is an advantage that’s hard to overcome.  When you are inside the decision loop of a competitor, it’s a huge tactical advantage.  Google got inside Facebook’s decision loop with Open Social.  Apple got inside the cell phone industry’s decision loop with the iPhone.  When you get inside the market’s decision loop, it is an enormous strategic advantage.  If you are lucky enough to be right, and lucky enough that a major market heads right to where your fighter plane’s guns are already aimed, you get the big success. 

Bill Gates got there with operating systems.  Was he lucky, yes, in the sense that Gary Kildall handed it to him, but he was also inside the decision cycle for a major market.  Gates saw personal computers as the wave of the future.  I can tell you from experience that dealing with IBM as a partner can be so painful as to appear pointless.  Yet Gates believe so strongly in his OODA conclusion and targets that he endured that pain where Kildall felt the world would beat a path to his door and it didn’t matter.  There are many of these critical OODA insights:

–  Jobs sees design as being the future versus raw technology.  Starting with the Apple II the has had a non-stop series of revolutionary design-leading products with just a few failures along the way.  He’s still at it today in a way that others are just not able to anticipate.  For digital industrial design, he has no peers.

–  Ellison saw the future of relational databases.

–  Bezos wanted to empower all retail to move online.  He would own the storefront, others would own the merchandise. 

–  Google owns the front door to the web in the form of search.

The list goes on, but these visions were important and enduring.  In some sense, these companies have understood their visions better than anyone else, and hence they dictated what the OODA loop would be and others had to follow.  Many of these guys are now a lot older, but they are no less entrepreneurs as their companies keep reinventing themselves through cycle after cycle.

Old age and treachery really can overcome youth and enthusiasm!

One of the first lessons I learned as an entrepreneur is that the loudest voices are from the edge.  In this case I was working on an early version of the software that would ultimately be bought by Borland and released as Quattro Pro.  QP was a spreadsheet originally concieved in Ye Olden Days before Windows.  It had a COWS (Character Oriented Windowing System) with full mouse support, pulldown menus, multiple windows, and all the rest.  Those loud voices I mentioned were from the owners of various offbeat video cards and printers that we had not yet announced support for.  Given the volume of their protestations, you’d think we had failed to deliver support for 90% of the world.  It was a real crisis, and so we rolled up our sleeves and started to try to see how many of these devices we could support.  The customer, after all, is always right, aren’t they?

You can probably see where this is going.  No matter how many of these cards and printers we supported, there were always more requests.  As far as we could see, each new card added at best a couple of new users.  It was a losing proposition.  So we backed off.  A lesson had been learned.  The loudest voices may not count for much.  In fact, the bleeding edge has a tendency to preselect loud voices.  Who else but someone bold and opinionated will choose to deviate from the mainstream path?  Who else will take the time to communicate so loudly with the vendor about their displeasures. 

The message here is not, “Ignore your customers.”  Rather, it’s to be suspicious of anecdotal evidence, and be aware that often the most vocal groups are really not very large.  It may still be valuable to support them from a PR standpoint, but don’t assume vocal is equivalent to large numbers.  Get some real market research done in order to find out.  These sorts of things are tragically knowable.

This is a mistake lots of Product Managers make.  They get upset when I ask them for real market research on exactly how many customers want a particular feature or platform.  Their view is that they’ve heard customes asking loudly, so there must be a lot.  It’s a mistake.  If you take it to the logical conclusion, you’ll spend all of your time building stuff for the Vocal Minority, and little will be done for the much more numerous and lucrative Silent Majority.  The latter mostly don’t complain loudly, they just take their business elsewhere if you don’t keep them satisfied. 

I was reminded of all this as I read a post from another blogger I follow.  Steve was upset because he had visited a retail site he wanted to buy from and couldn’t.  It seems their Rich UI makeover wasn’t working on his browser.  He couldn’t figure out why, and didn’t feel like it should be his problem.  His message was that if you have a web site you want 100% of people to get through no matter what, so Rich UI is a bad idea.  There is some wisdom in that, but it’s not clear whether it’s really the answer.  Presumably the Rich UI delivers a better User Experience and hence more sales.  That could easily more than offset a few unhappy campers, we don’t know for sure.

I don’t know what Steve was doing wrong, if he was doing anything wrong at all, or whether the site changed since he wrote the post, but I got on with my box stock IE7 and it worked great.  You have to figure this retailler didn’t just redo their whole site and fail to test it on the obvious combinations.  They might have, but the odds are better that the retailler tested it on a variety of common configurations, found it worked great and rolled it out only to find that certain much more rare combinations were not working.  Unfortunately, some of those combinations were in the hands of the Vocal Minority, so Steve wrote about it.

These sorts of things are hard decisions to make, but aiming for the lowest common denominator is rarely the winning strategy.  A fair number of great companies and products have been built by starting from products almost nobody could use because they were too far up the power curve and relying on the market to bring everyone up to capacity as the product grew.  Windows is a prime example of this (early versions were largely unusuable until we got Windows 3.1), but there are many others.

Figuring out the cutoff point is hard, but ulitmately you have to decide how many Loud Voices from the Edge you will ignore.