Friday, December 23, 2011

Gregory Slayton on Innovation

Dartmouth Tuck School of Business Adjunct Professor Gregory W. Slayton was in Beijing this month to give some talks about innovation.  Although creativity and innovation are of course important to me, I haven’t focused on this, partly because I think of it as an art (that you learn by doing) rather than a science (that you learn by studying). Obviously it's a little of both, and as a professor he’s studied it academically, so it was nice to get an overview of the state of the science.
He divides professional creativity into four components:
  • Product: what you make
  • Process: how you make it
  • Interpersonal: who you partner with
  • Strategic: future directions
Companies as well as people can be analyzed on these dimensions, and he asks us to rate ourselves (our “Creative Profile”) on each of the four components and assess where we want to be in five years.
He recommends three classic books on creativity:

as well as some readings I thought were useful about Pixar,  Alessis, and a highly-quoted HBR article by Teresa Ambile.

Other suggestions:
  • Keep a creative journal to write down new thoughts or ideas as they come to you
  • Appoint a “Creative Board of Directors”, mentors who will give you feedback on how to be more creatively successful
Gregory has an interesting background: Harvard MBA, McKinsey consultant, Silicon Valley businessman, and Consul General to the Bahamas. He's also had a longtime interest in Asia, and I enjoyed the short conversations we had during the breaks, talking about his thoughts on China and more.

Sunday, December 18, 2011

Zeo vs. the motion sensors

I don't have a Jawbone UP. I tried to buy one earlier this month but Best Buy wasn't stocking them until the kinks have been worked out. Several of my friends rave about it, though:  it's just a bracelet that you wear and it automatically measures your activity, including your sleep.

The UP is one of a number of devices that try to measure sleep based on your movement in bed. It has a tiny accelerometer that picks up the slightest little twitches of your hand while sleeping. Since sleep phases are often accompanied by such movement,  (the theory is) software can later stitch it together to guess how much time you were in REM sleep or deep sleep. But these twitches are just a proxy for the actual sleep phase,  so I'm skeptical that it can measure it as accurately as the Zeo, which uses special sensors to directly detect the electromagnetic activity in your brain. Still, maybe it's "close enough", especially if (like me) you don't have any particular sleep issues that need analysis to the nth degree.

Movement detection is pretty easy, and there are plenty of ways to do it.  There's even an App for that! Smart Alarm, by Arawella Corporation, cleverly uses the built-in accelerometer, plus the microphone, to measure your movement at night and guess the amount of various sleep phases.

How does it compare to my Zeo Sleep Manager?  Last night I tried both at the same time and here are the results:

myZeo Personal Sleep Coach



Smart Alarm Sleep

Answer: Zeo is way better. It’s not even clear that the motion-detection app gave useful information, and might even be outright wrong.

The motion detection method was wrong. It says I:

  • Slept one hour longer than I did. (9 hrs vs 8)
  • Had more phases of REM sleep (6 vs. 4)
  • Had less deep sleep (5 vs. 8+)

Look at the charts and you’ll see the difference.

I don’t know how well this compares to an UP, but I bet the motion-detection systems just aren’t very useful.  If you really want to measure sleep, I say get the Zeo.

By the way, my recommendations:  Zeo comes in two forms: an alarm clock version that doesn’t require anything extra; and a more portable, cheaper version that plugs into your smart phone. If you have an iPhone or an Android, the Mobile Sleep Manager is a little cheaper and smaller.  The Alarm Clock version is nicer if you don’t already have a nice bedside alarm clock, or if you don’t like sleeping near your phone.

Saturday, December 17, 2011

I'm sleeping better, but why?

I love my Zeo Sleep Manager, a cool device that measures my brain activity in order to help understand the quality of my sleep: the amount of refreshing REM versus Deep sleep, versus plain-old-ordinary light sleep, and of course the all-up total sleep each night. I started using it in mid-2010, and have used it to track my sleep most nights since then.

I don't have any particular sleep issues, though I was intrigued by the idea that I may be used to a particular quality of sleep, and maybe I just think I'm doing okay. Well, after using Zeo for this long I can't say that I feel all that different -- I feel fine, and always have -- but feelings are hard to measure objectively. Is there a way to measure my sleep quality more scientifically?

Zeo uses a measure they call "ZQ", which tries to combine a bunch of aspects of sleep into a single number you can compare across nights. After analyzing my data for the past eighteen months, I've noticed that my ZQ number seems to be going up over time, for no apparent reason.  At least, I don't think I've been changing anything about myself: I just go to bed when I'm tired, and wake up when it's morning. I haven't deliberately tried to change anything about myself.

But even without trying, my ZQ seems to have improved over the past year.  Since the amount of daylight varies throughout the year, I analyzed my numbers date by date.  (I don't have data for every single night, so my analysis skips the nights when I don't have a datapoint for both years).

Here's a chart that shows the difference in my ZQ compared between given dates in 2010 and 2011.

Sprague ZQ 2011 vs 2010

As you can see, there seems to be a clear trend of my ZQ improving, and for the past few months it's improved significantly – on the order of 20 or more points per night. I wonder why?

Sunday, December 11, 2011

Worse than a robot

I’m back in the US for a few days, a good time to order items that I can’t get in China, and that means borrowing the address of a friend to accept delivery. Since I come often, I try really hard not to wear out my welcome with these friends by making the pickup process as simple as possible.

This time I needed wanted a new iPhone 4S, which can have unpredictable arrival times, so I ordered it several weeks in advance. My friend happened to be out when the truck came, and I got an email notice offering to hold it instead at the main FedEx shipping center. Perfect! I thought: I’ll pick it up right after my plane arrives, early in the morning  with no need to intrude on my friend.

Seemed like a perfect plan until the lady at the FedEx counter asked for my ID. Of course I have my passport/drivers license, but the package was delivered to my friend’s name, and as the lady explained to me: “FedEx policy requires that the name on the package match the ID of the person picking it up”.

Well what I can do? I’m obviously who I say I am -- she sees my ID -- and I have the correct tracking number. It’s clearly my package. But policy is policy, according to the  counter lady. My only hope is to get my friend to call FedEx and change the name on the delivery. I explain that it’s early in the morning, my friend did me a favor by accepting delivery in the first place, and I don’t want to impose.  Sorry, she says.  “It’s policy”.

“Wait,” I say. “What’s to stop me from calling FedEx myself?”  I know the tracking number, the address of the original delivery, and I have an email that FedEx sent to me. Rather than ask my friend to call, why don’t I call myself, pretend to be the friend, and the problem is solved, right?

The counter lady hesitates. I have a good point, she admits, but now that she’s on to me, she says, she still won’t let me have the package because she’ll know it was me just faking to be my friend. The only thing I can do, she insists, is call my actual friend and get her to dial FedEx herself.

But that’s a hassle for my friend, who will have to drop what she’s doing to make a phone call, look up the tracking number, sit on hold.  I really don’t want to impose.  Too bad, says the counter lady. “Policy is policy”.

I called Apple. The person on the line was very friendly and accommodating, but Apple’s IT systems and FedEx IT systems are separate, so it could take as long as 24 hours before word of the different name on the address trickles into the FedEx office. The Apple person offers to speak directly to my FedEx lady, who replies “Nope: it’s policy”.

Finally, after too much time wasted already, I excused myself and went outside. I called FedEx and said I want to change the name on the delivery. No problem, they said. A few minutes later I went back to the counter lady, she looked up the entry and sure enough it’s okay to accept delivery from “Richard Sprague”.

Ugh.  What a waste.

This counter lady adds no value. By sticking so firmly to the rules, she was making herself into an automaton, the perfect job for a robot. Unlike a machine, though, she can’t work twenty four hours a day, and she needs to be paid.  So she’s actually worse than a robot!

If, on the other hand, she had used a little common sense -- the kind that is far more complicated to program into a robot -- she could have realized that my story makes complete sense. I am showing her a real ID, and I’m happy to give real additional contact information in case --against all logic--I am a criminal who somehow stole this tracking number, faked the email I showed her, and now is going through all the trouble of coming to the FedEx office -- in person -- to pick up a delivery of a brown box that has no indication of what’s even inside.

The US unemployment rate is too high, and there are a lot of proposals for how to “put America back to work”. But the unfortunate fact is that too many Americans are like this FedEx counter lady: doing work that is fundamentally replaceable by automation and robots. I don’t know what this particular women will be doing in five or ten years, but I know that if FedEx wants to continue controlling costs, they’ll need to look carefully at how much value she adds, and inevitably they will conclude that a robot is better for this work than she is.

It’s sad, because she, like all humans, has some skills that are extremely hard to replace with machines. But first she’ll need to start acting like a human, and not like a robot.

IMG 3843