Sunday, October 23, 2005

endurance sports and entreprenuerism

this afternoon between reading papers about locality-sensitive hashing, i took a brake and did one my standard weekend activities of going for a long run. for those of you who do not know me, i typically swim/bike/run six days a week and compete in triathalons and running races. people always ask why i train so much etc (even when i travel i am obsessive about exercising), and i typically answer 'relieves stress etc' - the common answers. but today, i came up with a better answer.

while running stairs in the stanford stadium with no else around, i realized that doing long workouts or races is very similar to starting a company:

1) intense discipline

when you start your own company, you report to yourself. you have to hold yourself accountable...no one will know if you do not finish your code or that proposal. you have to set milestones for the company to ensure progress. you have to build an operating plan

when training for a triathlon, you (typically) do not have a trainer that will get you out of bed to swim at 5:30, run with a headlamp in the winter and bike with you in the rain. also, you will often be alone on those workouts...you have to drive yourself. you need to build a training plan to prepare for the race and to ensure progress.

2) embrace the unknown

start-up companies are obviously high risk endeavors...and the outcome is never guaranteed. just working hard is not enough...you have to take risks. big ones. you may have to finance the company yourself...you may have to give up having a social life for months. all for an unknown outcome...but you believe

while racing, you never really know what is going to happen. when jumping out of the ferry boat during the escape from alcatraz triathlon, you just do not know what the conditions will really be like. how cold is the water? how big are the swells?

the unknown can be feared or leveraged...i say build on it.

3) exhaustion is a payoff

going home at 3am after solving a difficult computer science problem is thrilling...there is no other feeling. yes, you are tired. yes, you have only eaten crappy food for hours. but you accomplished it...

finishing a 6 hour race is...well...unbelievable. the human body can go much further that anyone ever really believes...

-> i am sure i can come up with more....just need some more sleep :)

Tuesday, August 16, 2005

homemade flamethrower

another good weekend project - a flame thrower. a definite must for halloween...

Monday, August 01, 2005

direct sales in enterprise software

as anyone that is involved in enterprise software is aware, the application of prior sales strategies do not work in today's world. very few examples of software still support the costs of operating a direct sales team (i.e. sap), so aa revolution in the enterprise software sales model is underway. however, several barriers exist:

resistance to change

many vp of sales that are recruited into enterprise software companies were directors of sales in the late 90s at companies that were *known* for strong sales. how many of those companies were riding the it spending wave of the 90s versus selling products of real value? i will let you answer that one - but regardless the generation of sales folks today were coached/mentored to ask for: high price point, thinking that a deal is never lost over cost, a sales rep and a sale engineer can close a deal. all of these tenets, in my experience, have evolved significantly in the past couple of years. deals do not have high asp (average sales prices), deals are lost over price and buyers are significantly more technical and demand a more technical sales relationship. yet, enterprise sales reps continue to try to use put a round peg into a square hole when it comes to evolving this strategy. people do not like to change a model that worked once....albeit in anamolous times.

gone are the middle it mgrs

one of the reasons for a more technical sale in 2005, is that post-bubble many middle level it managers were removed from organizations. they were replaced by highly technical architects that will only deal with people on the sales side that know what they are talking about. although a sale rep may try to navigate to the lob (line of business) owner to have a *business* discussion, the deal is still between this promoted architect and the technical sales rep from a compnay.

a solution?

one approach is to re-evaluate the entire licensing model of enterprise software today and move to a dual-license model that incorporates gpl/commercial license. on the direct side of sales, the license prices are lower and the margins are in the support and services provided (think mysql ab). the real opportunity then for enterprise software for new start-ups will be in providing commericial licenses to oem partners that can leverage their sales base and network for distribution. this is consistent and supported by the recent consolidation that is occurring in the market that is being driven by players like oracle. this notion of taking privately developed software and then moving it towards to the open source domain may be appropriately called 'private sourcing'.

just a thought, as the cogs (cost of good sold) for many public and private companies of software licenses for direct sales are all coming at an operating loss. thus, we need a change.

Wednesday, July 27, 2005

idea #312 open source search index...

i would first say that this may already exist, and if it does i have just not come across it....

not all open source projects are hosted on sourceforge and furthermore there are many academic projects with gpl/lgpl licenses that take a bit of clever googling to locate. thus, there may be an opportunity for an index of projects both in-development and completed for searching and identifying open source components in a more *simple* manner. this service will probably be more directory based than using a classic information retrieval approach and would of course be free :) while i applaud google's attempt at hosting another sourceforge like system, it will never address the above challenge...

Monday, July 25, 2005

work and life balance

recently brad feld (a vc) wrote about the life/balance challenge here and found it to be very well written. as someone who has endured a couple of start-ups and the effects that they have on your personal life, i have a similar view as brad in terms of viewing balance as a metric over a period of time versus a single day. there are days/weeks that i will go so focused on a specific activiy with no balance in other areas of my life...then i need to regroup and invest in the other areas. yes, i am only recently realizing this ;)

Saturday, July 23, 2005

etl for blogs

i got talked into starting another blog at yahoo360... and i am not frankly that into maintaining two dialogs. for those interested, it is here. it is much *simpler* to post photos via yahoo's blog, so as you will see there are most images from recent travels.

this multiple blog maintenance problem seems like a small oppt for some smart engineer with some time...

more thoughts on the migration of enterprise sw guys to the consumer space

so i have mused over the pros and cons of this migration, but recently an advisor pointed out an interested comment on this. he indicated that in the enterprise space, that typically when a company is started/founded that the *problem* is well-known and defined. as someone who falls in this category of having started an enterprise software company, i can certainly relate to this comment and agree with it. however, in the consumer space there is not the *problem* concept, but rather a company excels and expands based on offering a service that is innovative. the innovation may be in ease of use, cost or a product that makes life more fun and/or interesting. so can an enterprise entrepreneur start a successful company in the consumer space?

in boston's copley square, they offer *free* pads to sit in the grass area...and as you can see the level of theft seems to be at a minimum...california needs to take note of the *pad cart* concept Posted by Picasa

Thursday, July 07, 2005


this is a *poor* photo taken by a treo 650 of the roundabout at the Arc de Triomphe in paris last week. although obviously more advanced than katmandhu, the lack of coordination was certainly a similarity... Posted by Picasa