Thursday, April 07, 2005

Two Big Announcements

Tribe Hires New CEO...

I'm Moving To Typepad

Find both at my new swanky home on typepad...http://markpincus.typepad.com

Tuesday, April 05, 2005

Two New Investments

EVDB - the service, which is aimed at providing a more open way for people to find and publish all events, was launched at pc forum last month and recently backed by draper fisher. i think that evdb represents the right approach to a web service in that they aggregate all existing events from meetup (and hopefully tribe). then they enable people to tag them into whatever categories they want as well as publish their own. i particularly like the integration with ical which seems way underleveraged to date as an exciting way for groups to share calendars. i really dont get why evite hasnt turned themselves into a branded web service that could be found within friendster, myspace and tribe but that's another diatribe.

Merlin Securities - this new prime brokerage service was launched by my good friend aaron vermut with his dad, steve vermut. dad built bank of america's prime broker into a business that i understand was more profitable than starbucks. in addition to having great web interfaces and software, merlin represents a great opportunity to bet on the explosion of hedge funds, but more selling picks and shovels than digging.

Monday, April 04, 2005

Vertical Search Engines and Aggregators

last week, walt mossberg wrote about two new search services that focused deeply on a particular vertical segment. i wanted to share my own thoughts on this concept and the categories he highlights.

jobs - indeed.com - he points out that indeed is aggregating most big job sites. my first question is whether the most important driver for a job applicant is comprehensiveness. we hired a terrific office manager slash 10 other roles on craigslist. she said she chose craig because he offered 'the best' jobs. by this she meant the most relevant organic (created there) jobs for her search. in her case, this meant jobs from real people (hiring managers like me) and not from recruiters, or worse, staffing firms who she says will regularly bait and switch to harvest more resumes. i'd like to hear from people what the use case is for most comprehensive in jobs. seems to me that relevance (ie. not wasting time sifting through a lot of wrong results is more important).

if this search aggregation follows the path that the overall web serach took, the winner will be the 'google' of each space, meaning the company that shows me the 3 results that meet my criteria and not the 3000. anyone remember being impressed with alta vista that it searched the 'whole' web? anyone remember using alta vista to find anything quickly? no. it was a project.
Indeed's search results were often too vague. A search for "screenwriter" in California turned up postings for office assistants, movie producers and even someone to remove a computer virus -- but none for actual screenwriters.


people - ziggs - course i've been talking about the coming peopleweb and this sounds like one of the first aggregators out there, however, i dont get the business model. why would i pay $25 to be listed in their directory when i can get found on the web way more often in google? oh wait, they also will buy my name on google for $50 a year...lets see, if i got googled 100 times a week, and clicked on 10% of the time (which is high but i'm vain:) and it costs $0.10 per click, that would be $1 per month. whoa, high margin biz! i'd pay the folks at ziggs a lot just for the list of consumers who were that dumb and lazy! maybe they could call it the 'buy a bridge in brooklyn' list. also, how will anyone ever know to use this site? will it be so massively popular that a year from now we'll all be saying, 'hey, i'm going to go zigg that hottie'? if so, i'll eat my blog (and buy their customer list too:) maybe they should have called it 'shag.com'. at least i'd remember it and have more fun saying 'have you shagged me yet?'

and what's ironic is that people is the place you really do want comprehensiveness and relevance, but mossberg says...
And it seems to be aimed more at helping people and companies promote themselves than at providing comprehensive information. Ziggs search results provide only neutral or flattering biographies, not information that might warn you away from a person.

now here's what i dont get, when i 'zigged' myself it only shows me a 'mr mark pincus of pittsburgh'. if i google myself, i own the whole page. if this site is less comprehensive than google why would anyone ever use it?

one funny (to me) aside on this whole topic is that two yrs ago when i started playing with all this stuff, i bought my name on google (course now i could just pay $50 for it:). i described this as the ultimate web vanity plates. bonus is google tells you how many times you've been 'zigged', or i mean googled. jonathan abrams (who was found by ziggs but third on the list. wonder if he paid the $25?) outbid me on my name and put up an ad saying 'get the real scoop on pincus for $25). taught me a fast lesson about this new world where anyone can 'own' your name.

where will this all lead? how far are we from anti fan blogs where girls can 'out' players, employees can 'out' ruthless bosses? we already had fucked company. anyone want to launch fucked boyfriend or fucked boss or how about fucked vc? i'll sell you my domain, 'ratemyvc.com', which is a more polite version:)

all listings - oodle. this is a sight that didnt get featured by mossberg. it was just launched by craig donato and does a good job of aggregating all local listings for 3 cities to start with. my question is how will they get repeat traffic at a cost that allows them a profit with google ad sense? and what will happen when everyone becomes an aggregator, or worse, a standard emerges allowing all listings originators to publish their listings in whatever fashion they choose? what role will aggregators have when they have a disadvantage in traffic acquisition and no advantage in aggregating? this smart guy, jason dowdell, sees this as the beginning of a wave of vertical search players.
Dan Gilmore on The Wrong Standard...

That's also why I get frustrated by the anything-goes attitude on Wall Street and in Washington and so many other places where cynical people ignore the context in which capitalism operates. Corporations are creatures of the state, not God, and they exist in a real world where employees, suppliers and communities should be, but frequently are not, given sufficient consideration next to pursuit of profits.

A spectacular return on investment does not pretty much say it all. It tells only part of the story.


while i dont totally agree with dan's take, i agree with his sentiments that we need to evaluate corporations and more importantly their masters on more than just ROI. warren buffett has set an amazing example of how capitalists should behave and be rewarded. i'm crossing my fingers that he turns this latest chapter into another text book case of how an ethical capitalist should act.

Tuesday, March 29, 2005

Tribe gets ripped in WSJ today

my company, tribe.net, got a fairly negative mention in today's wsj (page B6). i want to make a few comments on the story. first, i think it's sad that we continue to maintain double standards, whereby it's ok to party like a freak in vegas (cause it stays there) but we're expected to maintain the pretense of all being clean 'white' people at work. frankly, i get numbed by how 'white' my world is in sf and often wonder how i can escape it and find more diversity. (i should say before getting shot by the pc police, that i mean 'white' as in we're all supposed to have non-tatoo'd bodies, wear dockers and generally look and act like good corporate cogs.)

ok, i'll be first to admit that tribe has a greater share of 'diverse' people than most networks, that it's not your grandma's internet, but common, this guy is saying he dinged a candidate for having tatoos and pictures without a shirt, and that he regularly 'friendsters' people to weed them out, that you dont want people to find out you're a big 'partier'? give me a break.

now, i do think that people have not yet awoken to the new reality of the peopleweb, where their public info will be findable by anyone. for that reason, tribe and most hosted community or profile networks will soon enable users to control what is shared and with whom.
Will the SEC ever get serious about making corporate insiders pay for fraud?

we are all watching yet another venerable old company (AIG) blow up over corporate accounting fraud and investor misrepresentation. same old story, investors lose billions, often everything, while corporate insiders say 'i'm sorry'. then the SEC tries to go after a small few with expensive criminal proceedings that rarely yield much. in this case the chairman, maurice greenberg, has decided to 'retire' and walk with $2.8 billion, in addition to the millions he's already sold. maybe there'll be a whistle blower. maybe in 5 yrs he'll face trial after an enron style bottom up investigation. this white collar crime has become the norm and the SEC is useless in correcting it.

here's the easy solution - MAKE THEM PAY!. it would be soo simple. just create a rule that says any insider who sold during a time of accounting fraud (or any restatement) has to disgorge all profits. this would immediately make every insider liable for the billions they made without bothering to prove any knowledge of wrong doings. the rationale is that they didnt deserve the profit, that they were selling a 'lemon'. california has this for cars, why cant the SEC have it for stocks? instead we punish all companies (especially small ones) with sarbanes oxley (also known as SUX), one of the dumbest laws our country has ever produced. it makes about as much sense as asking people to swear on the bible before testifying.

last year over 400 public companies restated earnings. think about the billions that would have been recaptured for the injured investors. think about AOL where the insiders walked with billions, virtually unscathed, while the poor saps who owned stock in time warner were left with a billion dollar tab to the SEC over fraudulent accounting. how about gary winnick who sold $600m at global crossing before it blew up? he chose to give back $30m to former employees. with this rule, he would have returned the whole boot. john moores sold $600m in his company, peregrine systems, before it too went bankrupt for fraudulent filings. maybe these guys didnt know about it. WHO CARES. just make them give back the ill-begotten gains. and the SEC can still go after any of these people in the current long drawn criminal process and nab one out of 400 or less at a huge tab to the taxpayers.

we're all disgusted with ken lay and the insanity they brewed at enron (see the enron documentary. you wont want to invest in another texas company!). maybe a few of these guys will see jail time, but even then they'll keep the loot.

so why doesnt the SEC enact this obvious rule? i dont know. maybe the same wall street guys who run the very stock exchange that regulates them have a hand here. maybe this would go too far in causing some real pain to a group that represents massive political donors.

btw, there is a clear precedent for this in the current 16B short swing profit rule that says if an insider buys and sells in a six month period he must disgorge the profit. i'm sure this was enacted after realization that criminal laws were ineffective against insider trading.

if anyone has ideas on how we push the SEC to enact real dollar regulations that will start to protect investors and not white wash the same old game, let me know.

Sunday, March 27, 2005

We Need OURMARKETPLACE.ORG

canter's ourmedia.org is a cool idea. i'm not sure i get how it will gain widespread adoption without 'network partners' to make use of it and bring the end users, but we'll see. my question is why isnt anyone developing 'our marketplace'? i mean an open source ebay that any user can use to buy, sell, trade, meet or do any activity without any friction; friction being fees and other commercial offers.

how this might work - the idea would be to create open servers and open databases that any other network could replicate and add onto - or that is, anyone could freely aggregate this db into their own and syndicate back into it.

benefits -

- consumers could see all items ranked by price or whatever measure they like.
- vendors could participate without having to pay for the marketing fees to get the consumers.
- consumers wouldnt have to ultimately pay the bill for being marketed to.
- networks could compete on value added services rather than bigger marketing budgets.

example - my friend works for the largest stamp vendor on ebay. they pay about $30k a month in fees (which just went up) and dont make any profit. one might argue that ebay brings him customers and only takes 10% and he can choose to leave it. however, makes me wonder...what if he could just tell the network of his available inventory and prices and the network can make sure any interested buyer immediately knows.

doesnt this sound familiar? it's called NASDAQ and NYSE. imagine if the stock market worked like the ecommerce on the web...maybe i want to sell 1000 shares of google, i can tell ebay or amazon or try 10 other networks that may or may not have an interested buyer. they may charge me varying levels of upfront and backend fees. the buyer then has to rummage through these different networks. because it's so fragmented neither side has a lot information what the last 1000 sold for or all the offers out there. neither buyer nor seller benefits, only the intermediary who looks to profit off of massive fragmentation.

if all the networks talked to each other in the same language this wouldnt happen. i could visit any stamp site and see all stamps and their prices historical and present. the marketplace function would move from customer acquisition and seller aggregation to value added services like surfacing most relevant offers or best reputation info or dynamic pricing.

i believe this can only be done in an open source manner as it must work like the DNS. it must be a freely distributed db and open protocal that any network can use and enhance.

if anyone knows of efforts that approximate this, i'd love to hear. and i'd say that my company, tribe.net, would happily participate in this type of endevour.