Archive for June, 2009

Complexity vs. Transparency

Wednesday, June 17th, 2009

What is transparency?  Even the most dogmatic laissez-faire advocates acknowledge it is necessary to efficient markets, but rarely define it. The usual definition is simply to make more information publicly available.  But this misses a fundamental observation: complexity diminishes transparency.

To appreciate this, consider an example we have all experienced.

In the past year, you have probably clicked through dozens of license agreements in various software programs and Web pages.  For example, on a Mac you clicked through more than a dozen agreements last year simply to run system software upgrades.

These are binding contracts.  But have you read all of them?  Any of them?

Of course not. There isn’t time. It could take 100 hours a year — about two weeks of work time — just to read your clickthrough agreements.  So you don’t.  Now you have a somewhat opaque business relationship with your software and Web service vendors, because the aggregate complexity of all such agreements exceeds your real-world capacity to absorb their meaning.

Similar examples include credit cards and insurance policies.  The typical homeowner has a half dozen of each, each with a binding agreement of 10,000 words or more, which is typically modified one or more times per year.  No one reads them, because no one can.

It turns out that the end buyers of mortgage-backed securities were in a similar position, but with a difference: they typically had a fiduciary duty to understand what they were signing.  Faced with the real-world impossibility of understanding the wildly complex MBS agreements — some of which are more than a thousand pages — they could have chosen not to buy.  Instead, they laid off the fiduciary responsibility to the rating agencies, simply assuming that AAA meant “safe,” as it had for the century prior to about 2004.

One could argue it was reasonable to believe the ratings, and that rating agencies blew it.  Or, one could claim (as rating agencies now do) that rating agencies were faced with the same problem:  the instruments are too complex to rate with confidence, and thus the buyer should fend for themselves.  But either way, the central problem is that the instruments themselves are so hard to understand that even professionals don’t get it.

This illustrates the complexity vs. transparency issue vividly.  By the traditional definition, MBS are utterly transparent, because you can obtain and read the entire agreement before buying.  But by any practical definition, they are opaque, because real-world buyers — even fiduciaries like bank investment managers — don’t have enough hours in the day, and possibly not enough theoretical background, to evaluate them.  Moreover, there is the well-documented psychological problem that almost everyone overestimates their understanding of complex things.

Caveat emptor?  The laissez-faire dogmatist might argue that those who don’t understand should simply not buy.  Again, true in theory, but again with a real-world, practical problem:  as MBS are currently prepared, if everyone who didn’t understand them didn’t buy, there would be no MBS industry.

“Well, maybe there should be no MBS, then!” some will say.  This is like saying, “fire is always bad because it burned me once.”  Fire is so useful that we tolerate occasional catastrophes.  Similarly, a stable MBS industry has such enormous benefits, from a capital efficiency standpoint, that we have an interest in seeing them presented simply enough to permit real-world transparency, and thus reliable ratings and comprehending buyers.  Tame the fire;  don’t extinguish it.

I conclude from this that federal regulation may want to focus on a “reasonable man” standard for simplicity in the presentation of financial instruments to certain types of buyer.  Not because careless buyers deserve it, but as a way to make complex instruments usable in the real world.

"Green Shoots" Pop Quiz

Tuesday, June 16th, 2009

The widely reported “imminent economic recovery” results mainly from:

  1. Record high credit card default rates
  2. Record high prime mortgage default rates
  3. 70-year low US factory capacity utilization
  4. 35-year high unemployment, still rising
  5. 65-year high fiscal deficit, accelerating
  6. Doubling of oil prices in 3 months
  7. Magical thinking

There is essentially no evidence of recovery.  Even a single quarter of positive GDP growth, or employment growth, or income growth.  Anything.  The market is up 35% on sheer hope.  Beware.

Offshoring retail

Tuesday, June 16th, 2009

Nostradoofus prediction for 2010:  Web retail is increasingly offshored.

To see why, pay a quick visit to the World’s Scariest E-commerce Site.  Savor the broken English, salted with nonsequitur French and Italian.  Admire the ragged formatting and non-matching fonts.

The site claims its location is “Alabama,” but they ask 20 business days for delivery, and the domain is registered in Shenzen, China.  Er, Alabama is far from California, but not that far.

If you actually buy something at that site, as I did, you’re greeted with the chilling purchase completion message, “Dummy string.  Actual purchase completion message goes here.” You then experience two weeks of deafening silence from “customer service” before your order arrives. But it does arrive.

What?  You’re not running out to buy from this site?  Before deciding, consider this:  on my first $100 purchase there, I saved 70% off the price of an equivalent item from a US site.  Including shipping costs.

So don’t look at the site itself, but rather what this site is trying to do.  Where is that $70 being saved?  They are cutting out the entire wholesale-retail chain, and shipping directly to your door from factories in Shenzen.  This is quite interesting.

And it gets better with scale. Once they are receiving 50 or 100 orders a day, they can ship them all in one box to the US, then break them up to trans-ship domestically.  Inventory? All in China, almost free to store.  Website?  Maintained in China, nearly free.

Perhaps most interesting is that, by locating near the factories, the retailer can eliminate inventory entirely, simply buying as needed directly from the factory.  This, in turn, permits the retailer to list a vastly larger catalog, since need not actually stock anything, but has immediate access to everything.

In the most extreme manifestation, the end buyer becomes the endpoint of a just-in-time delivery system, in which retail orders directly trigger production runs in China.

Yes, there are site quality issues. For now. But I would not bet against the ability of the industrial titans of Guangdong to solve that problem.  Web vendors of imported durable goods should be squirming, because this system is inherently and vastly more efficient.

Palm ruins another career

Thursday, June 11th, 2009

Ed Colligan today is officially out as CEO of Palm, ostensibly due to the fizzled Palm Pre product launch.  Perfect example of Warren Buffett’s aphorism that when a good manager runs a bad business, only the reputation of the business survives.

We routinely confuse great businesses with great leaders. For example, eBay, an unregulated natural monopoly, enjoys network effects so large that, after 1998, it would have wildly succeeded with a chimpanzee at the helm. Form that mental picture before voting in Ms. Whitman as governor.  Not to say she’s bad — maybe she’s great. But we have no evidence either way, because eBay’s strategic advantage renders good leadership almost immaterial.

As regards Palm, my opinion (as head of a top-5 Palm software developer during the peak of Palm’s popularity) is that Palm’s window of opportunity had already closed before Ed took command.

Colligan will do great in private equity.  It runs in the family:  his older brother Bud, also sharp, is a GP at  hoary VC firm Accel Partners. All longtime Palm folks who know Ed expect his best work is still ahead, now unencumbered by Palm’s limitations.

So now it’s Rubinstein’s turn to be tarred and/or feathered by Palm’s decline.  Judging from his Apple career, he is a brilliant product guy (caveat:  he’s the first Palm CEO that I’ve never met).  And Rubinstein’s own reasoning is obvious — no one can rise above VP at Apple. But Palm has none of Apple’s advantages: scale, the network effect of iTunes, retail distribution control, and the Apple brand.

Palm’s only realistic window was far earlier, ending in perhaps 2003.  At that point they had most of what Apple has now and a few things Apple doesn’t:  leading share in smartphones, multiple carrier relationships, and a huge, experienced software developer base, many times larger than that of any competitor.

That’s all gone now, and Palm is trying to scale a smooth teflon wall with its bare hands.  Maybe Rubinstein will find a career later in private equity.  It’s more forgiving, and intelligence is more predictably rewarded.

The calculus of joining a declining Palm

Palm’s executive headhunter approached me around 2004 as a potential VP of product marketing. Terrible fit, for a number of reasons on both sides. Two facets of that dialog are relevant here.

1.  Palm’s strategic options are hamstrung by its misplaced consumer ambitions

Our talks ended immediately after I proposed a product strategy:  abandon the hopeless quest for scale in consumer phones, and focus instead on ”enterprise” phones:  ruggedized, WiFi, barcode reader, high-resolution camera, and deep software API.  Acquire Symbol Technologies, the largest industrial PalmOS hardware vendor. Own the market for insurance claims adjusters, inventory clerks, security guards, etc.  Leverage Palm’s only hard-to-copy advantage — the world’s largest base of mobile software developers — with a more sophisticated OS and radically simplified software installation and update system.  (Apple did all of this a couple of years later.)

My position, then as now, was that Palm was already doomed as a consumer play, for lack of scale. Any ephemeral market share gain would be pyrrhic, because mass market consumer phones have very low margins. Motolora’s travails illustrate the problem.  Even for Apple, many times larger, with a strong brand and distribution advantages, success with the iPhone was by no means assured, and even now enjoys high margins mainly because of the network effect of iTunes.

2.  Promotion to First Officer — of the Titanic

From a career perspective, would it have been worthwhile for me to become VP of a public company in terminal decline (if offered, which it wasn’t)? Probably yes, but not at all clear.  So I can really appreciate Rubinstein’s calculus in joining and remaining at Palm.

Who knows?  Had I gone there, I might by now be settling into a comfortable semi-retirement in private equity…

This is my ninth message!

Thursday, June 4th, 2009

I know of a company whose main customer service problem is the occasional angry customer who claims he or she has already left several messages, though the company has actually received none.  The compay’s setup:

  1. Only one phone line.
  2. Only one person answers it and picks up VM.
  3. Same complaint has occurred with 3 different VM services.
  4. Company uses Google VM, which keeps a record of every call.

Thus, the customer’s claim seems as if it cannot possibly be true.

The callers all have one thing in common:  they are forgetful.  How do we know this?  Because when finally reached, they have all called for the same reason:  they do not recognize the company’s charge on their credit card bill.  The company reminds the customer, who says, “Ohhhh yeahhhh.”  Every time, at least so far.

Since these customers are forgetful about what charges they have committed to, perhaps they also forget that they have not left 4 messages?  My theory is that they may have called 4 times, but never actually left a message before.  Thus no contact was made, yet the customer has a vague feeling the company was unresponsive.

Suggestions invited from this blog’s massive readership.

Branding for Early Adoption

Thursday, June 4th, 2009

Early adopter behavior speaks volumes about the brand positions of Google and Microsoft.

With Google, the response to a new release is, “I will sign up for Google Wave immediately, lest I miss a developing trend.”  The response to a Microsoft release is, “I will delay trying Bing until proven, lest I waste time on yet another failed initiative.”

The Microsoft image develops in response to a series of failed or incoherent releases. Palm suffered the same image problem among its developers after a series of failed attempts to update its operating system after 2002.  Apple suffered the same problem during the 1990s, after repeated false claims of innovation.

The requirements for a positive brand image among early adopters are surprise and utility.  The new offering must not only work well and do something useful, but also must innovate in a direction unexpected even to industry enthusiasts.

So, for example, a product like Microsoft AdCenter is not only unsuccessful as a business, but also hurts the Microsoft brand among early adopters, because it is utterly derivative, does not work well (slow, poor browser compatibility, basic bugs in the site’s business logic), and is not useful (doesn’t generate enough traffic to be worth the hassle).

By contrast, Google develops its brand by innovating in unexpected, but useful, directions: big Javascript apps;  open APIs;  super-effective collaborative spam filtering.  Enthusiasts themselves can no doubt come up with more and better examples.

Underlying all this is the idea that early adopter branding is product-driven.  You actually have to have a good product.  You can’t promote your way to a brand in this arena.