The Consumer of the Future Will Be an Algorithm

Professional basketball player Jeremy Lin graduated from Harvard in 2010 with a degree in economics and a 3.1 GPA, but despite his great college basketball playing he went undrafted by the NBA that year. It was only after the New York Knicks lost two guards to injuries early in the 2011-12 season that they signed him at all. And of course, to the great surprise of the team’s management, Lin went on to electrify fans, scoring an average of 27 points a game in six consecutive wins.

But what I find fascinating about the “Linsanity” story is the fact that nearly two years before Lin started his first Knicks game there was one man, Ed Weiland, who predicted Lin’s success based on what his own unique algorithm said about Lin’s college playing statistics. In May, 2010 on his basketball stats blog Weiland wrote, “If he can get the passing thing down and handle the point, Jeremy Lin is a good enough player to start in the NBA and possibly star.” The amazing thing is that Weiland wasn’t a professional scout, just an avid basketball fan with a personal computer. He made his living driving a FedEx truck.

Many of us don’t appreciate just how empowering big data will soon be for ordinary people. We look at data as a corporate tool – “they” are tracking “us,” and we can only hope that the big corporations with all this data will use it humanely.

But big data can also directly benefit consumers themselves. Today’s big-company analytics programs are tomorrow’s smartphone apps. What will happen within a few years, almost inevitably, is that business algorithms will be making the offers, while consumer algorithms will be making the buying decisions.

We can already see the outline of this model beginning to take shape. Maybe you remember Farecast, the service that predicted when a particular airfare was likely to be going up or down, prior to flight. By drawing on 175 billion previous airfare data points and applying rigorous analytics, the service claimed to be able to predict when an air fare would go up or down in the next week, with an accuracy of more than 70%. Founded by computer scientist and U. of Washington professor Oren Etzioni, Farecast was acquired by Microsoft for $115 million in 2008 and its technology is now available through the Bing search engine.

After selling Farecast to Microsoft, Etzioni launched another price-forecasting service called Decide.com, which uses even more data and analytics to predict future price changes on a whole range of consumer products. Today, for a small subscription fee, you can access Decide.com on your computer or your smartphone, and find out whether the price on an item is likely to increase or decrease in the next two weeks or so. For instance,

• If you’re considering whether to buy an Apple iPad, Decide.com predicts that the $499 current price for the iPad Gen 4 will likely drop by $53 within the next two weeks (82% confidence). So you might want to wait a few days before you buy.
• Or if you’re thinking about an expensive Precor Elliptical Trainer for your home, be advised that the EFX 576i model, currently priced at $2999, is likely to increase in price by $233 within two weeks (75% confidence), so you’ll save money by buying now.
• Need an infant car seat? Decide.com predicts (with 96% confidence) that the $290 price for the highly rated Chicco KeyFit 30 will drop by $96 in the next two weeks.

And by the way, these are the predictions on Decide.com as of early Monday morning, July 8. But they change regularly, as new data come in.

You can set “alerts” on the items you’re thinking about buying at Decide.com, and you’ll get email notifications immediately when the price changes by more than 2.5%, or 5%, or 10%.

Decide.com also has useful and informative product reviews, helpfully summarized into percentile ranges based on both users and experts, so you’ll be able to see that the iPad Gen 4 is rated in the 95th percentile (very high quality) based on 5,194 user reviews and 10 expert reviews, while the Precor machine is rated in the 69th percentile based on 7 user reviews and no experts.

In a previous post (last year) I predicted the rise of “social filtering,” which basically means scanning the Internet for reviews and applying a higher level of trust to the opinions of your friends, or the friends of your friends. Facebook’s Graph Search was launched shortly after my prediction and is an early attempt at social filtering, while the Canadian startup Wajam.com is now a social-filtering aggregator, delivering your friends’ opinions on products, restaurants, web sites and other things from a variety of different forums.

So it isn’t really too hard to imagine the next step for a data-driven consumer shopping app, is it?

Imagine an app that taps into Decide.com’s price history analytics as well as Wajam.com’s social filtering information. Connect your credit card information to this app, preload the app with a list of purchases you’re considering, and authorize it to “pull the trigger” on a purchase when its algorithm calculates that the price is optimal, provided that none of your friends or your more distant social connections has weighed in with a highly negative review in the meantime.

My point is that your own shopping activities in the not-too-distant future will likely be sharpened with highly sophisticated data-and-analytics tools. And you won’t have to have a computer science degree either. You’ll have access to these tools even if your day job is driving a FedEx truck.

-------------

The story of Ed Weiland and Jeremy Lin is told in Christopher Steiner's excellent book Automate This: How Algorithms Came to Dominate Our World (2012)