Wednesday, November 09, 2005
Search is Not a Commodity Because Oil is Like Attention
Don't worry...it will all make sense in a few paragraphs :)
I came across Nick Carr's recent piece on search becoming a commodity today. As you'd expect from Nick, a cracking read - but ultimately, I think, no cigar.
Let's distinguish between two uses of the word commodity. The first is the way Nick intends it; the way economists sometimes use it - to refer to a pool of largely undifferentiated goods.
The second is the way it's used in the biz world - to mean a good which offers little return.
Now, there are times when the two overlap - when goods which are undifferentiated will earn few returns. But not always. And this is where Nick's argument falls down.
Nick is conflating the two meanings of "commodity" - arguing that since search is becoming undifferentiated, it will also necessarily become less valuable. Does this hold?
I don't think so. Let's think strategically for a sec to understand why.
Consider another example: oil. Oil is a commodity, in the sense that it's undifferentiated. But that hasn't stopped oil giants from reaping enormous windfall profits lately.
Why? Oil is scarce. But that doesn't equal profits. The reason there are big profits for a small number of players is that there are huge barriers to entry, which means that not everyone can afford to extract, refine, and produce oil. In the real world, that means oil is not just scarce, but it's a concentrated industry, where, when there's a demand shock (read: China), a limited number of players divide a big pie.
Now, search is kind of similar. Like oil, markets set prices for search: the price of a keyword, or a chunk of attention, is set by auction mechanisms, on, for example, AdWords. Like oil, Google benefits disproportionately from a demand shock (read: ad money flowing massively to the www) because advertisers bid up keyword prices.
And crucially, like oil, attention is becoming increasingly scarce at the margin.
But more importantly, it's also a capital and knowledge intensive business. It costs a great deal to turn inputs into outputs. This creates strong entry barriers. That the industry is concentrated - there are only a handful of players, each with significant market share - is a testament to this.
So even if search is a commodity - that is, players are becoming less and less differentiated - it by no means necessarily follows that it will be devalued. The structure of the industry tells us that quite the opposite should be true.
Of course, the litmus test is reality: Search makes a 30%+ (and growing) margin. Barring some kind of discontinuity, I think it's safe to say it won't erode anytime soon.
What this means is intuitive: returns to GOOG/YHOO/etc won't likely fall until entry barriers fall. When they do, search might be commoditized, in the sense that returns will fall. If so, however, we will certainly see the structure of the industry change drastically, and concentration will decrease.
That said, I do somewhat take his point about relatively low switching costs. In fact, we just discussed it to death last week...scroll down.
Short version: switching costs are relatively low for consumers, so by building scope and scale, incumbents are trying to maintain market power. At the limit, completing "indexing all the world's information" makes switching costs irrelevant, and means you will probably only search at Google.