Yahoo and the economists

Yahoo is hiring academic economists and other reseachers, we learned today.

This development is driven by a couple of things.  As a newcomer, Google stole a march on incumbent Yahoo and Yahoo would like it back.  Google has data miners, but it appears often to act as if it can win and sustain advantage with great technology alone.  If Yahoo could add new user sensitivity, this might help close the gap. 

Second, Yahoo has staggering amounts of data that can be used in house and sold to others.  In July, there were 482 million unique visitors, each of whom, over the course of the month, spent around 4 hours viewing around 240 pages.  This generates data like crazy: over 12 terabytes a day. 

It’s a little like the old story from the Dole pineapple plantations of in 20th century Hawaii.  Dole would can the pineapples and pour the juice into the ocean.  Finally, someone said, "You know, I think you could sell that stuff."  In this case, of course, the "juice" doesn’t just make the corporation richer, it makes it smarter.

In fact, this should work like a virtuous circle.  Orphan data get a home.  Academics get grist for the mill.  Yahoo gets more consumer centric.  More consumers join Yahoo and spend more time there.  More data is generated. The cycle begins again. 

But why these academics? Economists have formidable powers of pattern recognition to be sure. And they are good at working with just this kind of data.  By why proceed is if the numeric measures were the only ones available to us? Economists too often act as if they have taken a vow of noncontact, that they cannot engage people unless their existence has been registered at a distance, and they cannot know about these people unless data have been metriculated out of the real world into the realm of the number. 

Well, if we are dealing with the census data for a couple of years ago, these are all practices well advised.  But that 12 terabytes that Yahoo will collect today is from people who are not only alive, and in fact probably still on line.  If we think we see a pattern, there is no reason why, with the appropriate permissions, Yahoo could not poll players and convene focus groups on line. Indeed, there is no reason why the economist could not make contact with a living, breathing, blogging individual and determine the "whys" that make the "whats" make sense.

The trouble of course is that economists don’t ask because they have the suspicion they already know.  The actor is self-interested, rational, and advantage seeking. But come on.  Reason not the need, as Lear is made to say, or our lives are "cheap as beasts."  If this were the only motive at work in the human communities, markets and cultures would be simpler, flatter, more predictable and less "exuberant" than they prove so persistently to be. 

We can discover these other motives, extracting them from those terabytes of data, but not if the Yahoo economists think them away. 

My conclusion (and I think you saw it coming; it is, after all, self interested) is that all that the mountain of Yahoo data is a task, a trek for economists and anthropologists working together.  Take us along as sherpa if you have to.  But  take us along. 


Delaney, Kevin J.  2006.  Hoping to overtake its rivals, Yahoo stocks up on academics.  Wall Street Journal.  August 25, 2006.

4 thoughts on “Yahoo and the economists”

  1. Agree with you completely, Grant!

    I have probably said this before on your blog: I believe marketing exists to the extent that the assumptions of neoclassical (textbook) economics are false. People and companies are not individual utility maximizers; they are not even would-be individual utility maximizers; they discount future events in a manner inconsistent with (neoclassical) economic theory; they do not have perfect knowledge, not even about the past; they do not have unlimited information processing powers or memories; nor are there even such things as commodities.

    Of course, the standard defence of these assumptions is that in modeling economic phenomena we have to start somewhere. There are two responses to this claim: First, why here, when these assumptions are so obviously false? And second, how do we know we can get from these absurd assumptions to somewhere close to where we want to be, where the assumptions are more realistic? We can only do that if the relationship between the underlying phenomena and our model of it is continuous; in other words, if it is the case that a small change in our assumptions still keeps our model close to the phenomena we are modeling. Every business and marketing experience, and every statistical instinct I have, tells me this is not the case with markets.

    Marketplaces are complex, adaptive, reflective systems, not machines; and models of markets need to use metaphors from biology and complex systems, rather than on 19th century mechanics.

    Of course, the real answer to “Why here?” is because these unrealistic assumptions make the math easier. When economics ceases to be a discipline which puts mathematical tractability ahead of insight into human motivation and behaviour they might get somewhere.

  2. As an anthropology major in college I sympathize with the sentiment of this post, but as an IT professional in a major corporation I have to say that the social sciences have almost zero credibility among techies. The fact that Yahoo is willing to hire even economists is pretty amazing. The sad fact is that the reputation for post-modernism that anthropology and similar fields have built over the last few decades in academia has overshadowed among many people in business – and especially IT – whatever else they may have to offer. The ability to quantify results and at least pretend to be a “real” science has allowed economists to by and large escape this. At least at my company, if you can’t show ROI your project has no chance, and I’m pretty sure if I tried to get approval for an IT project involving anthropologists I’d get laughed out of the project approval board meeting. Sad, but true, and you can thank academic anthropologists for the reputation (or lack thereof).

  3. The phrase “user sensitivity” is key. Yahoo seems to feel that users of their free services are sheep — the services are second tier and tend “tilted” so that the resources go towards their favor. Meanwhile, Google’s win-win philosophy seems to shine through in just about everything they do. Yahoo still has their place, but they need to sure up their marketing position.

    -Steve (

  4. Obviously, I disagree with Peter’s broad brush. Demand curves really do WORK; when you’re trying to figure out what will happen to fuel consumption if the gas tax goes up or which buyers to give discounts to, quantitative approaches using market data are pretty effective. And they are furlongs better than focus groups or ethnographies, which would be like hammers used to drive screws in such problem contexts.

    Economists’ allergy to asking people things directly comes from their fear that people will say things that don’t reflect their true information/motives. Introspection is not entirely accurate, even if accurate may lead to embarrassing conclusions, and even if accurate and not embarrassing requires effort to conduct and communicate. Economists like to observe peoples’ behavior under changing incentives and constraints. Even the heterodox behavioral economists focus much more on what people do than on what they say they do.

    Grant is right that this allergy leaves a lot of running room for anthropologists to make a contribution. Sometimes what people say is really helpful in understanding their decision process. And sometimes, where hard-to-objectively measure factors such as meaning are involved, only a cultural analysis can put meat on the economists’ analytical skeleton.

Comments are closed.