The Limits of Measurement

I had a conversation with a business partner the other day about data, and the importance of being data-driven.  There’s little question that data is important – as we’re fond of saying around the office, it’s critical to avoid “the land of the thinks.”  But to this person, data was everything – as he put it, if something couldn’t be predicted and measured, his business wouldn’t do it.


This strikes me as a fundamentally limited way to think about decision-making.  While having data is critical, it’s vital to keep in mind that data is a tool – and like all tools, it has its limitations.  How?  Here are the three primary problems I see with being too beholden to data:

  • The “round peg, square hole” problem.  One of the realities of business is that decisions must often be made in low-data environments.  You can test, gather intelligence, do research, but ultimately you may be guessing:  how a market will react, what actions regulators will take, how a counter-party will respond.  It can be comforting in such situations to rely on whatever information is at hand, and elevate its importance.  The problem is that what you may consider “data” may well be little more than “anecdote.”  Or worse yet, modeling based on all sorts of rickety assumptions – “garbage in, garbage out” elevated to the status of solid forecast via the numbing comfort of a financial model.  This can obviously lead to overconfidence and an unwillingness to adapt as reality starts to deviate from the model.  Far better, then, to identify such decisions as what they are: educated guesses.
  • The elevation of irrelevancy.  The benefit of having measurable goals is obvious:  by focusing on those items, and holding people accountable to those goals, you greatly increase the likelihood that they will be met.  Which is great, as long as those are the only goals that matter.  The problem is that there may be other things that matter more.  These may be new opportunities or risks that pop up, or simply more amorphous things that are enormously important but which are hard to measure or set goals for.  The more you get down in the weeds with measurement and goals, the likelier you are to drive grocery clerk behavior: ticking off boxes on the to-do list rather than asking, every day, “what is the best thing I can be doing for the business?”
  • Paralysis.  After nearly 20 years in business, I’m convinced that the ability to embrace ambiguity and make confident decisions in the absence of complete information is the single biggest driver of success.  Yet “analysis paralysis” runs rampant throughout the workplace, as people are deathly afraid of making a decision without having as much data as possible (often defined as “more”).  The problem, of course, is that there is a cost to not acting, of not making a decision.  And you often can’t wait for more data – either because it’s simply not going to be available, or because by waiting on it the window of opportunity will close.  A business that worships at the altar of data is going to find itself constantly at risk of letting hand-wringing and delay keep it from pushing forward.

Ultimately, data can be so comforting, but too often what it provides is a false sense of security: the idea that because we are measuring something, and because we are making progress against that measured goal, we are succeeding.  But businesses – especially those that are new or growing rapidly – operate in an environment of ambiguity and constant change.  Our checklists, goals and to-dos can only offer so much help on that front.

Can Sheriffs Defy the Feds?

Ah, my home state of Oregon.  Reacting to President Obama’s push for gun control, a number of county sheriffs in the Beaver State have published letters stating that they will not help federal authorities enforce gun control laws they deem unconstitutional and – more ominously – they they would not “stand idly by” while the G-Men interfere with the constitutional rights of law-abiding Oregonians.

At first blush, it doesn’t seem that county sheriffs – elected officials all – should be in the position of determining the constitutionally of the laws they are employed to enforce.  But consider: prosecutors can choose not to defend on appeal laws that they do not believe are constitutional.  Notable recent examples include USAG Eric Holder’s refusal to defend the Defense of Marriage Act, and California’s refusal to defend Proposition 8.

What’s more, there’s also precedent for county sheriffs refusing to be the instrumentality of federal law enforcement – and in the gun control arena, no less.  In Printz v. United States 521 US 898 (1997), the Supreme Court held that provisions of the Brady Bill requiring local law enforcement to carry out background checks on handgun purchasers were unconstitutional under the principle of dual sovereignty.

So are the sheriffs in the right?  Not so fast.  First of all, they’re likely reacting to the potential for bans on assault weapons and high-capacity magazines.  While such laws may well be unconstitutional, sheriffs won’t be administering those regulations.  What’s more, the promise to not stand idly by while federal authorities enforce the laws sets up the potential for a serious jurisdictional crisis.

We remain a nation of laws, and if you see a common thread in the examples of prosecutorial discretion or local exercise of sovereignty, it’s that they all happened within the judicial process – not a face-off between state and federal law enforcement.  If these sheriffs want to make plain that they will not be the tools of the Feds, the time to do so is when actual laws are in place, and within the bounds of the legal process.  It’s not by releasing vague, inflammatory letters.

Car2Go – First Look Review

Car2Go Seattle

Last month, I wrote about the upcoming Seattle launch of car-sharing service Car2Go.  It’s now hit the streets, with a fleet of distinctive blue-and-white Smart cars flung about town.  So of course, I tried it out (and if you live in Seattle, you should too – you can sign up now and get the initial $35 fee waived, along with 30 free minutes of driving, using the promo code “SOUND”).

The Good

The core concept of the Car2Go service is a little bit amazing. It’s like ZipCar, but without the designated parking spots or need for all trips to be round trips.  Which means that if, say, you ride your bike to work but need a different way to get home, you can pick up a car downtown, drive it home and simply leave it in any public parking spot within the Car2Go home area (Car2Go has a deal with the city, covering all parking, even in pay and zoned parking areas).

I approached the service with a fair bit of skepticism, assuming the home area would be a tightly-defined area in Seattle’s more densely-populated areas.  I didn’t expect it would cover my residential corner of Capitol Hill, but figured I could walk from the edges of the home area.  So I was surprised to see that the home area is almost all of Seattle (and 100% of Capitol Hill).  I could leave the car right in front of my house!  Hell, I could leave one in the leafiest corner of Washington Park.  That made the service far more interesting to me.

The cars are all Smart cars.  I hadn’t driven one before, but it’s kinda fun.  Like a riding lawn mower with more pick-up, or a go-cart.  But despite the diminutive size, the car has a surprising amount of headroom, and room in the back that could probably hold four bags of groceries.  I wouldn’t want to drive one down I-5 to Portland, but it’s perfectly fine for tooling around town.

Finding cars is easy, using either the Car2Go website or the mobile app.  You can grab any available car you happen upon, or reserve one up to 30 minutes in advance.
Once at the car, it unlocks similar to a ZipCar, using your Car2Go prox card.  Unlike ZipCar, there’s an interactive touch screen and pin required before motoring away.  Once the trip begins, you’re charged $.38/minute, with an hourly max of around $14.  More expensive than ZipCar, sure – but the fact that I could grab a car downtown and leave it behind parked in front of my house was huge.

Or it would be, if I could have left it in front of my house.  And this is where we get to the not-so-good.  Car2Go has a few bugs to be worked out the system.

The Bad

For starters, the website.  It looks like something you would have encountered in a Barcelona internet cafe, circa 2002.  Car2Go is a European company that has only just recently made its entry into the US market, and it shows.  Let’s just say that the US remains several steps ahead of Europe when it comes to website UX.  The Car2Go site struggles with geolocation, and is fairly difficult to navigate.  But I’m sure they’ll improve, and it’s really a minor annoyance.

Ending a trip, however . . .

One of the really appealing things about transport services like ZipCar and Uber is how smoothly the process of ending a trip goes.  Everything happens in the background.  With ZipCar, hold the prox card over the external reader in the windshield for two seconds and you’re done.  With Uber, collect your bags, thank your driver and you’re off.  Or Capitol Bike Share – find a station, park the bike, look for the green light and that’s it.

It may seem a small thing – and lord knows this is a first world problem – but it’s a major part of the user interaction design that these services remove nearly all of the transactional friction at the end of trip.  This is also one of the appealing things about mass transit – just step off the bus.

With Car2Go, you’ve got to deal with the interactive touch screen again to end your trip.  And it needs to acquire a cellular signal and confirm back to you, presumably to ensure the car is parked properly and to confirm the trip is ended.  But for me, this process took upward of a minute, and – far worse – required my trying 3 different parking locations before I found one that could acquire a wireless connection.  That’s a big problem, particularly when users are paying by the minute for short trips.  It probably added 7 minutes onto what was a 15 minute trip.  That’s an unacceptable level of friction.

Now, I may have gotten a bad car, or perhaps Car2Go is using a small number of proprietary wireless transmitters.  But it’s an issue they’ll need to clean up – perhaps by using GPS to confirm that a trip ends in the home area.

(**UPDATE:  I did have a balky unit.  I’ve since used Car2Go a half-dozen or so times and the “end trip” process has been smooth, hassle-free and hasn’t taken more than 5-10 seconds)

That said, Car2Go is certainly a welcome entrant into the Seattle transport scene, and further reduces the chances that we’ll become a multi-car family again anytime soon (at least until my kids start driving).  Now, bring on Seattle bike share!

The Low Relative Cost of Bike Repairs

I’m horrible at staying on top of bike maintenance.  It’s one of the reasons I moved to single-speed, and why I can have three bikes and still lose a day of riding due to mechanical issues.

I try to do a lot of my own maintenance and adjustments.  I can change a tire, and I’m pretty handy at swapping out pedals or wheels.  But I’m crap with the brakes (and if I had derailleurs, I’d be completely lost).  And I don’t clean my chain nearly often enough.

So when I finally get around to dropping my bike off at the shop for some much-needed deferred maintenance, I always find myself wondering why I hadn’t done it sooner.  The ride is so much better (and in the case of my latest repair, safer), and it’s peanuts compared to car repairs.  The latest work I had done was the automotive equivalent of a complete brake rebuild and front end alignment.  Cost, including parts, labor, and tax?  Less than $70.

(shout-out to The Bicycle Repair Shop.  It’s not the most convenient shop for my commute, but worth it – great guys and terrific service.  And a sweet dog.)