The Two Best Things I Saw Today

Swarm Light, by rAndom International...

'Swarm Light' by rAndom International / Design Miami / Art Basel 2010 from rAndom International on Vimeo.

From the Vimeo page:

The ‘Swarm Light‘ is an experimental light installation with a real ‚collective consciousness’ that subtly reacts to the viewer’s audible presence. The installation is a contemporary example of how the arbitrary boundaries of fine and decorative art, design and utility are no longer of immediate aesthetic relevance. An apparently inanimate object, ‘Swarm‘ unites crucial aspects of rAndom international‘s continued experimentation with light, behavioural responses and interactive spatial environments.

Cit/ID: Honour Your City

A great collection of illustrations in which type, inspired by various cities, is the star.

Both Sides of the Mouth

One of the aspects of banking reform that gets occasional, but not enough, airtime (IMO) is fees, and the effect of fees on the effective rate of return consumers get on their accounts. One of my pet favorites is overdraft fees. When I was but a young buck, overdrafts were terrible, awful things... events which would earn the transgressor an imaginary scarlet "O" tattooed onto their forehead. Now, well... you tell me. Overdraft has become a form of expensive credit, because instead of refusing to honor your check or approve your debit card transaction, the bank will allow it to go through and charge a hefty flat-rate fee (my bank is $35.00 / overdraft).

When consumer groups claim that always going into overdraft rather than having the option of having transactions declined, the refrain from bank spokespeople is that "our customers appreciate the convenience of knowing their transaction will always go through." Really? I'd love to see how that survey was worded. When politicians complain about the size of the fees, the boilerplate answer is that, "the fee has to be large enough to dissuade people from going into overdraft."

Wow. My head spins. Let's count the reasons:

1. The two arguments are obviously at odds with one another. They want to provide a service, yet dissuade their customers from using it. More likely, they see an opportunity to take advantage of an instance of information assymetry, and the past stigma of transgression is their justification for current gouging. To be honest, at this point I would pay for the service and convenience of not having my transactions honored if they would cause an overdraft.
2. The penalty rapidly loses teeth relative to the size of the crime. The fee is charged for every transaction, however small (or, in my experience, large). This means that the effective interest rate on the "loans" is much lower for larger transgressions than for small. The more I spend that I don't have, the less I have to pay for each dollar of it. The word for that is "regressive," and it affects lower-income folks (coincidentally those most likely to be affected by information assymetries) the most. This sets the stage for exploitation.

I made two purchases last week that resulted in overdraft fees due to my dumb mistake. The first was for $182.90, which resulted in a fee of $35.00. The second was for $16.93, which resulted in the same fee. That impiles effective annualized interest rates of 230% and 2480%, respectively.

I think the word for that is usury.

Photo: Paper Lanterns 2

Paper Lanterns 2, originally uploaded by Chango Perezoso.

Here's another that turned out well. This was a bit more intentional, and as I shot digital (Canon SD-870IS), the gratification was much more immediate. This was taken in the Presidio in San Francisco, on 25 Dec 2008.

Photo: Conversation

Conversation, originally uploaded by Chango Perezoso.

My wife regards plant photos as facile, and despite the fact that I've often given into the temptation of instant gratification and cheap thrill that comes with shooting a pretty flower or leaf, I tend to agree with her in principle. Every now and then, however, I get one I'm proud of. It's almost always by chance. This one I shot back in 2001 or so, in San Francisco.

How to Scale a Bakery (?)

My wife told me a story the other day that made me sad and curious. The woman who made our wedding cake had a bakery in San Francisco's Potrero Hill neighborhood. We visited on a number of occasions; it was a small storefront, but the pastries were delicious and the bakery was always busy.

A few months ago, the bakery was closed. How could it have closed? It produced splendid product, was in a busy neighborhood retail strip with plenty of foot traffic, and little or no serious competition. Certainly it didn't suffer for lack of business.

Indeed, business was fine. In fact, it was too good. Read that again: the owner closed the bakery because business was too good.

This fascinates me. According to my wife, the baker closed down because she (1) couldn't find help she could rely upon, and (2) had no time to do what she loved, which was bake. Instead, she had to spend her time managing her employees. I brought this up with my Organization Architecture professor, and he exclaimed, "YES! BAKERIES! Bakeries can't scale! I don't know why, but they can't!" He proceeded to tell me about a similar problem faced by one of his friends, who is a baker as well.

What stands in the way of a bakery scaling? While products from one high-end bakery to the next are somewhat differentiated, they're differentiated mostly by the creativity of the bakers. If there are fifteen high-end bakeries in a given area, why couldn't there be instead five bakeries, each three times the size? What/where are the scale economies that bakers can exploit?

Perhaps most importantly, however, is the question of the appropriate organizational and incentive structure for a bakery. In the end, each pastry is a little work of art, not a product squirted onto a conveyor belt. An example of a successful bakery is Arizmendi Bakery in San Francisco, Emeryville, Oakland and Berkeley. There is a notable difference between Arizmendi and our friend's former shop: Arizmendi is a co-op, meaning that the workers are also the owners. Arizmendi, then, is more a guild of artisans than a company of employees. Perhaps this is the secret?

Information and the Crisis

This is a classless cross-post from Marginal Revolution, which linked to the original article. This is something I've been suspicious about for a long time. Various culprit have been taken to task for the crisis... the instruments (securitized mortgages, credit default swaps, etc.), the players (non-creditworthy borrowers, greedy mortgage brokers, greedy insurance salespeople), and the market system itself. But in theory, this all should have worked out. In fact, buying insurance on mortgage bundles should have been almost unnecessary, as theoretically securitization diversifies a mortgage portfolio thereby reducing risk of catastrophic loss.

Theory though, it seems, always assumes high-quality, timely, and understandable information. This is not what we had, and there was no incentive to make it thus. Certainly there were incentive problems (as well as vetting problems, as many of the less ethical mortgage brokers turned out to have had criminal records) along the way. But shouldn't this have come out in the perfect-information wash along with everything else, and factored into the market price of said securities?

Hopefully there's a silver lining in this: some way that the quality, complexity, and timeliness of information can be taken into account in pricing various derivative securities (or any securities). Certainly we shouldn't do away with the instruments (many of them, anyway). There has to be some way, however crude, to derive a crude factor based upon information quality, and use it to normalize the value of a security relative to alternatives.

Customer Service I Like

We've all had to deal with customer service nightmares. One of my "favorite" stories is of an insurance claim I filed a while back. The adjuster was a bit odd; I never met him in person, but whenever I spoke with him over the phone I felt as though I was being interrogated by a Smith. "Missster... Middleton, yes. How... are... You... today. We sseeem to have an... iss-ue." The issue was finally resolved, one year after I filed my claim and not to my satisfaction. The adjuster stalled; he never answered his phone, even if I called him 30 seconds after he tried to call me; and hethrew up every roadblock imaginable (and a few that are still mysterious). The claims officer I worked with at the insurance company was apologetic, with the refrain, "I understand. It's a big company."

While this was going on, I had a lot of time to think about the meaning of the excuse, "it's a big company." If a company is too big to provide the product that its customers have paid for, then doesn't this make it "too big a company," in the sense that it's reached a diseconomy of scale? It seems to me that in that apology lurks an admission that the company is inefficient, that people shouldn't buy its products, and that its shareholders should sell. I won't name the company, but it seems that my armchair analysis was correct, because they have since pulled out of some sectors of the California insurance market.

I'm not writing to complain about poor customer service and corporate inefficiency, however. I'm writing to praise what I think is the best customer service experience I've had — ever — with a software company. The company is OmniGroup, and the product is OmniPlan.

My previous post discusses the company and mentions the product. OmniPlan is project planning done correctly: it doesn't look and feel like a tarted-up spreadsheet. The interface is clean and extremely easy to understand. Error messages are concise and helpful, often providing suggestions which can be acted upon simply by clicking a link (contrast this with Project's tendency to render paragraphs of text in modal dialogs, making any good suggestions). While it can't yet form "links" between projects like Microsoft Project can (I'd love to be able to have a master rollup project and resource pool, and multiple detail projects owned by different editors linked into the master), for students and for most professionals organizations, it's perfect.

Again, though, I'm here to talk about excellent customer service, not excellent applications. A new version of OmniPlan, one which supports OSX 10.5 (Leopard) is in beta; I've been using OmniPlan 1.5.2b2 for some time now. It works extremely well, and has some nice new features such as filtering which I find very handy. However, once I built up a complex project with around 250-300 tasks, many with multiple resources assigned at various allocations and with inter-ask and inter-task-group dependencies, I noticed a slowdown and some reproducible crashes.

A word about being a beta tester (or being a software user of any kind). Software has bugs. I don't care how well crafted it is, or on what operating system it runs; if it's much more complicated than "Hello World," there are going to be issues. Good engineers and good companies acknowledge this and make a good effort to minimize bugs and address them promptly. This is actually what "beta" software is all about; once software becomes mature enough to use, it's actually better in the long run to release it into a hostile environment so that bugs which don't show up in testing can be caught and fixed. If one chooses to run beta software, one enters into a sort of agreement with the software developer. Part of that agreement is to report issues as they arise. This means that issues bugs or crashes should be reported, to the extent possible, along with the steps to reproduce the problem. Developers need to know about the events that led up to a problem, or the context in which it happened, and only the beta user can provide these details.

It might not seem like this is conducive to being lazy. However, I'm concerned with long-term laziness; call it "Sigma-Laziness" or "Net Laziness" if you like. I like software that makes my life better, not just software that makes each five minute stint of my life better. OmniPlan, like all of Omni's products, makes my life better. I was therefore happy to drill down into a problem I spotted over the weekend which caused a crash with the beta I was using, and to provide as much detail as I could in the seemingly anonymous automated crash report and feedback system that's built in to Omni apps.

Within a day, an Omni "Support Ninja" contacted me, and asked me for more details. Some I was able to provide, some I had lost. In the end, it was confirmed that the details I provided made it possible for the developers to reproduce the problem I found, and ensure that it's fixed in the third beta, expected shortly. All of this was delivered with very rapid turnaround, and the Support Ninjas seemed genuinely concerned, thankful, patient and happy to work with me to eek out the details. While it took some extra time on Sunday to articulate the problems I was having, in the long run, issues are being fixed and my life will be better.

Meanwhile, Omni has shown itself an example of a company that can be successful, produce great products, and stand behind them — no excuses. I don't know much about their internal organization, but if I were to guess I'd say that they have a lean organization, a culture of ownership and accountability, and a great hiring process. They also seem to really like what they do. They may be right at their particular economy of scale. I hope they stay that way, and that other companies take notice.

And no, I'm not on the Omni payroll. I just like to give credit where credit is due.