Monthly Archives: May 2007

man vs. machine – the definition of search 2.0

Posted by Todd Sawicki on May 30, 2007 View Comments

Those who know me know one of the guiding prinicples in my business and product thinking is that automation is king.  I believe that smart automation will always beat out manual processes when automation can be applies.  The thesis has certainly lead to my involement in a series of startups these past few months.

And so it’s interesting to see my fellow silicon alley alum jason calacanis (yes I date back to the early, early, early days of silicon alley – 95-98 and yet I know I’m an old man by startup standards per fred wilson) use a lens from the wayback machine to rollout a manually edited search engine.  To an extent that’s where Yahoo started with a editorially developed internet directory and what google took over with their automated search engine. 

I have no idea whether Mahalo will overcome the challenges of having a manually edited search engine, but I will give him the benefit of the doubt given the success of the weblogs’ coterie of blogs.  Effectively Jason’s applying the same logic from that experience to search – target high traffic concepts with slick editorial and walloh users will come.  It will be interesting to see if it works.  Also what I don’t know yet is what automation and tools will Jason and Co. build in to reduce the burden on the editors.  I am sure more to come on that front.

What’s interesting is that Spock, the search engine for people that as disclosed previously I have a previous affiliation with, is building what I believe to be the most sophisticated automated cataloger of people ever built.  And yet even then, the Spock team knows there will be a limit in its ability to be 100% accurate with regards to the results and identifying too data points as the same person.  Therefore Spock will rely on its users to refine and improve the results.

Maybe the approaches are that different in the end as manually edited with a side of automation vs. automation with a side of manually editing sound like they will likey meet in the middle.  Maybe that’s what search 2.0 really is – not so much vertical search but community improved search.

stumbleupon bought by ebay

Posted by Todd Sawicki on May 30, 2007 View Comments

Announced earlier today – eBay bought Stumbleupon.  I ran across Stumbleupon back in 2004 and ran the toolbar for awhile but ultimately dumped it.  Never got into the user generated recommendations thing myself.  That being said, I will say Stumbleupon has been this blog’s biggest driver of new users so it’s definitely got some juice as perhaps the Digg for the long tail.

Also some folks have wondered whether this means eBay’s looking to get into the recommendations space and given my involvement with Loomia I am curious to see myself.  Personally I don’t think an application based approach that relies on human voting is the right approach vs. Loomia’s automated system that relies on some very sophisticated analytical algorithms in order to recommend products or content.  But then again its not clear eBay plans to use Stumbleupon for user recommendations, but it will be great to see what happens. 

What is clear is that helping users find the products and content they are looking for whether through a lean forward approach like search or a lean back approach like Loomia’s recommendation service is something of growing interest for web publishers and retailers.

people search: worth $30 billion?

Posted by Todd Sawicki on May 26, 2007 View Comments

Don Dodget does a great analysis this morning on what 1% of the search market is worth.  Putting my Spock hat on (disclosure – I did some work for them last winter), you can see why the market is excited about people search.  Assuming that people search in its broadest definition represents 30% of the search volume, then Spock’s playing in a market that could be worth $30 billion.  That’s a big market.  And it’s easy to see what David Stern & Bill Quigley from Clearstone Ventures and Ken Elefant from Opus Capital saw when they put in $8 million into Spock last fall.  And as a shareholder of Spock – I sure hope Don Dodge is right!

spock invitation

Posted by Todd Sawicki on May 25, 2007 View Comments

I have one invite left for folks looking to get into the private beta for Spock - the search engine for people.  Post a comment if you’re interested and I’lll forward you the invite. 

Spock Search of the day:

valleywag and micropayments (been there done that)

Posted by Todd Sawicki on May 24, 2007 View Comments

Valleywag has an essay today pitching hard for a future of micropayments.  Would the madness stop.  As many, many, many folks have discussed many, many, many times micropayments are doomed to failure.  Having a lot of experience in consumer marketing online, I can confirm the theory that the biggest drop in sales is when you go from free to 1 cent.  That hit on users that imposing even a 1 cent price or tax creates has doomed every micropayment service to ever hit the startup landscape (there should be an entire section for these companies in the proverbial startup graveyard).

The good news is in all of this is that there already is a $17 billion dollar micropayments business online.  What didn’t I just say that micropayments are DOA?  Yep, but there is an alternative – advertising.  Advertising is by FAR the best solution for monetizing micropayment level transactions.  Advertising can easily cover a penny for something price.  That’s why the web 2.0 universe is free.  It’s not free – it’s just being paid for by advertisers who are more then willing to pay you way.  Is this a new model?  Nope, broadcast TV has been using the same model for 60 years.

technorati re-launches – but their search still lacking

Posted by Todd Sawicki on May 24, 2007 View Comments

Discussed widely online (such as at read/write web), Technorati recently upgraded/re-launched their site to expand their search focus beyond just blog posts.  That might be all fine and good, except Technorati isn’t that good at search.  Like all good bloggers, I do the ego-check at Technorati to check on how many blogs are linked to this blog.  And consistently they miss probably half the links to this site (interestingly enough Google’s blog search product is able to find the other half that technorati misses).  Not sure what Technorati is using as their crawler (either an in-house spider or perhaps a processing of Alexa’s web-wide crawl) but whatever they are doing is falling well short of their promise.  It’s great that Technorati put a new coat of paint on their house, but unless they can truly be the place to find blog links and posts, well they won’t make it.  Hopefully with the new coat of paint dry, their engineers can now focus on fixing the leaky plumbing so to speak.

google’s new tag line: we do nepotism

Posted by Todd Sawicki on May 23, 2007 View Comments

Henry Blodgett (yes that Henry Blodgett) posted a good critique of the google investment in sergei brin’s wife biotech startup.  And to his point – what in heaven’s name is google thinking?  What a total PR blunder.  No matter how google’s kingpins try to justify this investment, the public will never buy it.  It screams nepotism and in this case like most public matters – perception is reality.  Now there’s nothing illegal about it, but if I were a google investor I would start to wonder who’s minding the store and whether or not sergei has jumped the shark.

cycling detour: landis hearing

Posted by Todd Sawicki on May 22, 2007 View Comments

Found a great blog – trust but verify - that is doing liveblogging from the landis hearing.  I have no idea whether or not landis doped in france, but reading the today’s and yesterday’s testimony sure makes the french lab accusing him of doping look like a bunch of keystone cops.  Pathetic.  Anyway, here’s a quote that caught my eye:

Q (landis atty): concusion about their ability and competence to operate the instruments.

A (

davis landis expert witness): they (lndd operators) clearly did not understand the instrument. I had to help them load the software on the machine [referring to testing of b sample with

davis as landis’s rep observing test]. They were obviously trying to help each other during the reprocessing and did not generally know how the software worked.

Wow.

Microsoft-AQuantive Followup

Posted by Todd Sawicki on May 21, 2007 View Comments

A quick followup to my post on the MSFT-AQNT deal announced last week.  There has been a lot of consternation in the commentary about the deal price across the internet perhaps highlighted by this post over at GigaOm.  First I want to clarify my earlier post – I wasn’t saying MSFT was crazy for paying the EBITDA multiple that they did for AQNT.  I was just noting it’s high price tag (and since I am currently on the sell side of the equation given the size of startups I am involved with I am a big cheerleader for high multiples these days).  Whether it was a good strategic decision or not, only time will tell (and given my perception of MSFT’s track record in the space I am on the skeptical side of the fence).

Based on my own M&A experience which involves being directly involved with the decision to buy a dozen or so Internet-related companies, if the deal works out basically no matter what was paid, then it will look a bargain in the end.  And if the deal fails, then no matter what you paid, then it will have been too much. 

the value of twitter

Posted by Todd Sawicki on May 18, 2007 View Comments

Jason Calacanis (someone who i met often back in my silicon alley days even if he wouldn’t be able to remember me from a hole in the ground) posts about his desire for a twitter pro account.  Since the rise of twitter I’ve been researching a number of startup ideas in the sms space and one of the big stumbling blocks is the cost of sending and receiving sms messages.   SMS pricing reminds me of buying bandwidth and hosting back in the 90′s as in insanely high and prohibitive for all but well funded startups – you’re looking at 2-3 cents per msg (sent OR recieved).  Supposedly some of the SMS aggregators as they are called (the guys who handle bulk sms on behalf of the carriers – essentially SMS ISP’s) are now offering flat rate packages in the ten’s of thousands of dollars per month but that’s still cost prohibitive today.

Jason estimates he’s on pace to send 5 million twitter posts/msgs a year (numbers add up quickly don’t they!).  At $.02 per message that would be $100,000!  Jason’s offer of $250 is a bargain for him and likely a terrible deal for Twitter if/when they actually figure out to monetize their sms traffic.  Now I would love to see a $250/yr plan myself because then some of the ideas I am looking into would make great financial sense.