Thursday, February 17, 2005
Economies of Speed
Economies of speed are kind of a maligned subject in the econ lit. Chandler introduced the notion to account for the advantage that high throughput created in the industrial era (via high capacity utilization leading to lower fixed cost intensity).
What's interesting is that little meaningful work has been done on the concept since Chandler. This is curious, given the fact that speed is considered a crucial variable by the digerati in predicting the success of many digital (and non-digital) businesses.
First, let's critically distinguish the notion of speed economies from first-mover advantage. FMA can be a result of many variables (learning, network fx, etc). Economies of speed, in the Chandlerian formulation, are fundamentally about the speed of production (throughput, how fast a good moves from input to output) - not time-to-market (which refers to how fast you can enter a given market).
Let's attempt a bit of formulation. Economies of speed are important in industries with naturally low entry barriers and high switching costs. In this situation, speed becomes a crucial variable. Even if supplier and buyer power are relatively high, speed gives you a mechanism to co-opt an entire market and exert upstream and downstream monopoly and monopsony power. The 'economy' in economies of speed is that because switching costs are high, faster throughput (think faster tagging rates in del.icio.us vs Flickr) raises your competitors' relative customer acquisition costs, creating a massive cost disadvantage even if their products have the same price/quality ratio as yours.
Where do we see markets like this? Well, leaving aside the obvious example of ecommerce apps, with the advent of Web 2.0, the entire media industry is about to become driven by speed economies, because Web 2.0 creates exactly the conditions described above: low entry barriers (ie amateur media/distributed economies of scale) and high switching costs (ie across various aggregators/distributors).
That's a pretty important implication, and it tells media industries to launch experiments in massive parallel so they can search the Media 2.0 landscape fast, and quickly co-opt entire Media 2.0 markets by discovering and turbocharging the dominant distribution model first. In other words, the implication of speed economies is Media 2.0 plays who can also leverage first-mover advantage will be able to exert massively asymmetrical market power vis a vis their competitors - there will be serious winner-take-all dynamics for Media 2.0 publishers/distributors.
In fact, we can think of iTunes as the first example of just how important speed economies will become - iTunes current market dominance is largely due to it's relative throughput advantage: signing more content faster than competitors at a better price/value ratio.
So, I think there are two strategic questions if you have a (potential) speed advantage: how do I lower entry barriers in my industry, and how, beyond speed economies, do I raise switching costs? If you can do these succesfully, and you've got a speed advantage, you'll crush Media 2.0 markets.
Begs the question - (1900-1970) distributors --> (1970-2000) aggregators --> (2001+) ?? What are the properties of aggregators 2.0?