Friday, December 26, 2008

Loosely-Coupled Economies

Yesterday my good friend Mike Curtis and I were talking about the current economic condition and how our shared experience with technology might inform our thinking about it.

Anyone who's been paying attention to (Internet) technology over the last 15 years has a good intuitive sense for the value of loosely-coupled systems. Loose couplings provide a measure of elasticity that can dampen the impact of a shock to any one part of the system. The tradeoff for this increased flexibility is decreased efficiency: tightly-coupled systems are typically more optimized for the specific task at hand and as a result incur less overhead. So by comparison, loosely-coupled systems are less efficient.

My suggestion was that as the global economy becomes more interconnected, it will need to also become more loosely-coupled in order to avoid the kind of cascading failures we're seeing now. But what about the accompanying decrease in efficiency? Well, I'm not so sure there needs to be one.

A more loosely-coupled economy would be one comprised of entities that are smaller and more numerous, and interact with each other at arm's length: ecosystems instead of conglomerates, freelancers instead of employees, many-to-many instead of one-to-one. But this system can only come into being if transaction costs between such entities becomes low enough. Frankly, I think we're getting there.

At my company, we outsource a lot of our non-core work including payment processing, server hosting, and even marketing. Our employees and contractors provide their own computers, phones, and offices. Virtually all of the corporation's connections—both to other companies and to people—are what I would call loose couplings. And technology makes the cost of the company's transactions with those loosely-coupled entities incredibly efficient.

It's simply cheaper for us to use PayPal to accept credit card payments than to process them ourselves. We're satisfied with their service but could switch to Google Checkout or Amazon Payments if we needed to. Our marketing function draws upon the expertise and creativity of tens of thousands of people around the world, but is completely automated and is designed in a such a way that it always pays for itself. And moving our server operations from one supplier to another can be done in a single day. We could move all of these functions in-house but doing so would not only make our business more tightly-coupled (and therefore more susceptible to the failure of any of of them) but also cost us more for every transaction.

I believe the dichotomy between tightly-coupled, highly-efiicient systems and loosely-coupled, less-efficient ones is false. In my experience, loosely-coupled, highly-efficient businesses are absolutely possible. And I believe that model shows the way forward for the broader economy—but there's a hitch.

Charles Fitzgerald, Microsoft's former GM of Platform Strategy, wrote about the difficulty of transitioning a company's business model from low-volume, high-cost to high-volume, low-cost:

The problem is to get to high volume, low cost, there is usually a low volume, low cost waypoint. The radical change in cost structure required is just too painful for most companies to contemplate.

lo-lo-detour.gif I believe there's a similar waypoint in the transition I'm proposing. For companies (or economies) to get to loosely-coupled, highly-efficient, they have to go through a loosely-coupled, less-efficient phase (illustrated here) which isn't an appetizing prospect, especially in times of economic hardship. But a push from government could help.

There are several things government could do to support and accelerate this transformation: make it easier and cheaper to form and dissolve corporations and execute contracts electronically, make it possible for individual small investors to put money into private companies, and maybe most importantly, guarantee people the freedom to create and own their own intellectual property even while employed by another person or corporation. The bizarre and retrograde IP serfdom that is imposed as a matter of course in Silicon Valley is in my opinion this country's greatest hindrance to innovation. I'll be writing much more about that in the future.

So, what are your thoughts about my suggestion of a loosely-coupled economic future?

No comments: