Showing posts with label velocity. Show all posts
Showing posts with label velocity. Show all posts

Wednesday, May 11, 2011

A 10X change in VC business model

 TechCrunch reports about continuing expansion of Y Combinator, a VC firm that in direct competition with angel investors targets very early stage startups:

Y Combinator just gets bigger and bigger. They invest in batches, twice a year. The last batch had 44 startups, an all time high. But that record is about to be shattered – they’ve accepted 60+ startups for the Summer 2011 batch soon.

Why this is a 10X change? Three key reasons:

1. The size of each deal is 10 times smaller than a typical VC deal in the valley.
2. The number of startups funded is more than 10 times greater than in a typical VC firm of that size.
3. The portfolio turnover envisioned (6 months) is about 10 times faster than the 5-year exit strategy horizon prevalent in the industry.

Acceptance rate is 2.5-3.5%.

tags: innovation, 10x, velocity, business, model,  startup

Monday, April 05, 2010

Trade-off of the day: Commitment vs Costs

Commitment to a service is a good deal for the service provider, but bad for consumers:

...prepaid customers [month-to-month vs 1-yr+ contract] generate less revenue per month than post-paid customers. They are more likely to switch back and forth between service providers. And they typically use more network resources, averaging around 2,000 voice minutes per month compared to about 1,000 minutes per month, than post-paid customers, according to Craig Moffett, a Sanford Bernstein equities analyst.

Commitment to a regular long-term service reduces the hassle of looking for the best deal and setting up a new account. And you end up paying for the convenience through over-priced plans.

tags: trade-off, communication, network, problem, solution, infrastructure, velocity

Wednesday, August 12, 2009

A 10,000X change

The pattern of progress in genome sequencing looks like Moore's law on steroids:

Aug. 10 (Bloomberg) -- Helicos Biosciences Corp.’s gene- sequencing machine mapped the genome of a Stanford University professor, who developed the technology, for less than $50,000 using the labor of just three people, researchers reported.
The cost has dropped from about $300 million required to decode a full human genome during the early stages of gene- mapping, when multiple computers and machines were needed plus the labor of more than 250 scientists and technicians

Add to it the ability to turn on an off specific gene expressions and you get a possibility of human evolution within one generation.

Can you imagine a world in which people mutate faster than viruses?

Tuesday, December 09, 2008

Innovation in e-tailing

Over the last week ( 12/2 - 12/9) I've been tracking Amazon's bestsellers in the HDTV category. To my surprise, I discovered that prices for most popular TV sets sold and shipped by the e-tailer fluctuate over time. For example, the price of Samsung's LN46A650 HDTV varies from $1,499.99 to $1,614 ( see the graph below).



I wonder if they run some kind of a reverse auction to maximize revenue from the holiday traffic.