Wow. Sometimes I feel incredibly fortunate, and last night was one of those times. I was up late, in an energetic classroom, judging final projects for Neema Moraveji and BJ Fogg. They were dual instructors of a Stanford class on Calming Technologies. For those of you who don’t know much about BJ’s Lab, it’s worth learning more. The New York Times recently profiled one of the Lab’s classes in an article aptly called “The Class That Built Apps, and Fortunes”.

I was blown away by all of the students. They were thoughtful and compelling. They were quick to learn from their mistakes. And most important for me, they each held a clear vision of the social benefit that would result from their work. I’ve been fortunate to have taught a number of classes, and never before had I seen such consistent competence and interest in a course.

Then it struck me: the students were great, but the common factor to all of them was BJ and his team. BJ has cracked the code on how to teach students effective app design. His app design methodology flows readily into a solid foundation for a business. Whether they realize it or not, what these students are also learning are the fundamentals of customer-centric, purpose-driven entrepreneurship.

I hadn’t thought that this dimension of entrepreneurship could be taught. Nonetheless, here I was, watching student after student display the right fundamentals on business iteration. I felt like I was watching the Silicon Valley version of “Stand and Deliver”. Jaime Escalante showed that it was possible to make AP Calculus approachable to East Los Angeles high school students. BJ was proving to me that successful product design and entrepreneurship can also be realizable -- given clever teaching, commitment, good listening skills, and hard work.

An amazing message. I suspect they’ll have a lot of inspiring material to work with.

Sent from my iPhone

Congratulations, team, for making Business Week’s “Best Young Tech Entrepreneurs 2010” list!  Well deserved!

One of the reasons startups fail is that they don’t actually listen to their customer.  It’s a lot easier to craft a product in one’s own mind and push it out, rather than cherish who the customer is and meet his or her needs precisely.

I sometimes like to suggest to entrepreneurs that they visualize their customers as their children.  I say this because I want to encourage the entrepreneur to feel a deep degree of care and curiosity. Each customer has his or her own unique personality and needs.  It’s the entrepreneur’s job to figure out how these needs overlap and whether they can be satisfied well enough by the product being offered. 

Just like in this video, the content is out there.  If you cherish the customer, you’ll be able to hear the answers.

I think Volkwagen’s PunchDub campaign is a great example of how to embed game mechanics into a customer awareness/acquisition campaign.  For more on how to see game mechanics everywhere, check out Amy Jo Kim’s presentation from Dave McClure’s Startup2Startup:

It’s possible to think of ways to make most parts of the consumer experience fun, including things like registration.  So make it a challenge to see if you can apply know game mechanics to the parts of the experience that seem tedious.   

Bummer.  System dynamics are always a tricky thing to understand.  Take this recent study, for example, written about in the New York Times.  Apparently, antioxidants such as vitamin C and E inhibit the impact exercise has on preventing diabetes.  The possible reason?  Exercise produces low levels of oxidative damage, which may be the signal to the body to become more insulin sensitive. 

In short, most feedback systems have self correcting mechanisms to restore equilibrium.  Otherwise, they would sustain as systems.  The question I like to ask is what causes a system to tip?  And how do you know?  So has twitter usage tipped?  And will foreclosure rates tip and accelerate in worsening as unemployment gets worse? 

Who knows?  But I’m going to stop taking my vitamin C supplements.

Looks like initial jobless claims have hit a peak.  The four week average is now 20,000 less than the high of 658,750, set four weeks ago.  Some economists believe a drop of 20,000 or more is an early indicator for the end of the recession. 

Let’s hope Seamus Smyth at Goldman Sachs is right!

This random analysis is for a new parent friend who at one time thought life would get easier after his wife gave birth.  Well surprise! 

Since I usually have way too much time on my hands and use it to collect random data and do random analysis, I thought I’d share some of the most obsessive data work I’ve ever done.  I decided to graph each of my kid’s sleeping patterns.  The chart below is for my third child.  Each day for the 12 weeks of his life, I noted (okay, I convinced my wife to note) every time the baby woke up or went to sleep.  The graph below plots out his sleeping record during his first 3 months of life.  On the chart below, the white areas represent when the baby was awake and the grey and black areas represent when he slept.  The black bars represent the first major sleep after 9 pm.  Each rectangle represents a 15 minute increment. 

Things to note for this child: 1) sleep initially is quite random but begins to clump together; 2) nighttime sleep remains quite irregular, which is what all parents feel and complain about; and 3) it’s crazy that someone would bother to collect all of this data and then plot it out.  Oh well.  As you can see, by Week 11, I threw in the towel (or rather my wife stopped putting up with my obsession). 

Plotted another way, the chart below graphs the length of nighttime sleep for our third child compared against his older twin siblings.  Also on that chart is his weight and average sized feeding each day.  After doing this analysis, I was surprised to see that his average feeding size didn’t change once he was a month old.   

Well, I better get back to more random analysis.  No more kids are on the way.

“In the week ending April 4, the advance figure for seasonally adjusted initial claims was 654,000, a decrease of 20,000 from the previous week’s revised figure of 674,000.”

I love it when apples are compared to oranges and people make a big fuss out of it.  The Weekly Claims Report has great data, but it’s very easy to be misled by what it says.  The problem is that the report compares an amount that’s preliminary, called “Advance Figure”, against a revised figure from the prior week’s amount.  What we really want to know is the revised figure from a given week against the previous week’s revised figure. 

When you go through the process to determine how these “Advance Figure” numbers get adjusted, the answer is simple: they get adjusted upwards.  From the period of the beginning of 2008 to today, these figures were adjusted on average by 2,000 additional claims, with a median adjustment of 3,000 additional claims.  They’re adjusted upwards roughly 6 out of every 7 times.

Why this matters is that readers and the press that cover these reports can get misled in two ways.  First, the initial claims number is consistently lower than what it becomes, making the claims number look marginally better.   Second, that preliminary number is compared against the prior week’s number which has been revised upwards, thereby making it appear that the change in claims is improving more than it is. 

To highlight this impact, take a look at what the 4/2/09 press release said:

“In the week ending March 28, the advance figure for seasonally adjusted initial claims was 669,000, an increase of 12,000 from the previous week’s revised figure of 657,000.”

What it would say today would be the following:

“In the week ending March 28, the revised figure for seasonally adjusted initial claims was 674,000, an increase of 17,000 from the previous week’s revised figure of 657,000”, or

“In the week ending March 28, the advance figure for seasonally adjusted initial claims was 669,000, an increase of 17,000 from the previous week’s advance figure of 652,000.”

These might seem like small numbers, but in short, the finalized change in claims turned out to be 42% higher.  The previous week turned out to be 63% higher.  If you go through the last 68 press releases, which I did!, you’d find that the sum of the changes reported in these releases missed one out of every two additional initially weekly claims.  That matters.

Plotting the revised numbers, and adjusting the April 4^th number upwards by the average under-reporting in 2009, you get the graph below.  I don’t have much color to add here other than even a 20,000 decline in initial claims doesn’t look like much.  To me, I don’t see a permanently flattening yet in the curve.  Claims don’t appear to be accelerating, which is good, but I suspect that the number will continue to climb for at least several more months.

Job loss appears to be decelerating. Even though BLS numbers are revised on average almost three times, the rate of misforecast is declining from missing one out of every two jobs lost, early in 2008 to now only missing one out of every 5 jobs lost. So adjusting for the expected undercounting of job loss in the preliminary March 2009 number, it seems to me that it will be finalized to exceed the January 2009 number, but not by much. Job loss for now looks like it is decelerating. That is good news.