Support Marketplace

How Google searches are leading economic indicators

Autocomplete predicts what you're looking for based on previous online searches.

Kai Ryssdal: Think about all the Google searches you've done today. Multiply that by 365 days in a year, multiply that by pretty much everybody around you, and what you have is piles and piles of data. Data that's valuable not only for advertisers, but for economists as well who are mining it for all kinds of things they can learn about us. Which is a convenient way to introduce Sendhil Mullainathan. He's a behavioral economist at Harvard. Sendhill, good to talk to you.

Sendhill Mullainathan: Nice talking to you too, Kai.

Ryssdal: So, what's up?

Mullainathan: I have a question for you Kai.

Ryssdal: Of course you just met me but that's all right.

Mullainathan: That's how I start all my conversation. Imagine you're stuck in a bunker deep under the ground. All you have access to is a little sliver of the Internet. The only thing you can really do is see what other people are doing on it. So, what I want you to think about it is: Could you possibly learn anything?

Ryssdal: So here's what I can learn: I could learn about the American election, probably, and I could learn about Kardashian and Paris Hilton and all that stuff.

Mullainathan: That would be a good topic to learn something about. It turns out you can learn a bit more that. If you just followed how many people were searching for the words "monster.com," if you saw a big spike up, it's time to celebrate -- albeit by yourself in a miserable bunker but -- it's the New Year. And you can get pictures into what people are sort of thinking about others. Just go onto Google and type "Why are Indians" and then look for the autocomplete.

Ryssdal: Oh, that's horrible! So it autocompletes this way and I'm talking -- obviously -- to an Indian guy. "Why are Indians so rude? Why are Indians so smart? And why are Indians..."

Mullainathan: Hey, actually that one's pretty good.

Ryssdal: "So rude, so smart, so cheap" and then the fourth one is "Why are Indians vegetarians?"

Mullainathan: Exactly. If you start to look closely at phrases like "foreclosures" or "pawnshops," it becomes a leading economic indicator. So when these things start to go up, well it looks like the economy is starting to tank. You know, we really want to know which way the economy is going and we want to know it before it goes there. So, can we look and see if whether people are searching for things like car purchases? 'Cause those are pretty good leading indicators that indicate people are starting to feel times are good.

Ryssdal: And have them truly be leading indicators.

Mullainathan: Exactly. Once the sales date comes in from Honda...

Ryssdal: It's too late.

Mullainathan: ...it's too late. And so if we want to know how Nashville doing, that's kind of hard from a data point of view. But there are a lot of people in Nashville searching for a lot of things. So from an Internet data point of view, it's not that hard. Here's an interesting way to collect data that's pretty useful, but you're getting it in a totally new way.

Ryssdal: Is there a way -- I mean since we're on Google and it controls the world -- is there a way that you can do this without using the search function? I mean it's not just search terms, can't be.

Mullainathan: No, it's not just search terms. When you search, you also get these ads. How is it that this guy ended up at the top of the list? Well, he paid.

Ryssdal: Right.

Mullainathan: And you can ask, how much does it cost? So if you look at a word like "savings," to end up at the top of that list -- it's like a $1.

Ryssdal: So that is: the companies that want to be at the top of that list have to pay Google a $1 every time somebody clicks on it.

Mullainathan: Exactly. Now for the words "Payday loans," it costs about $10. So you're already getting a sense of where the money is. Helping you save? Eh... somewhat profitable. Helping you get a payday loan? Very profitable. But the highest one I ever found: $50.

Ryssdal: $50 per click?

Mullainathan: Per click.

Ryssdal: Oh my gosh!

Mullainathan: If someone is searching for the term: "structured settlements."

Ryssdal: You have to explain.

Mullainathan: OK, so now you get to learn something.

Ryssdal: And I'm not even paying you $50 a click.

Mullainathan: So, when you go and get -- let's say -- a lawsuit and the company pays you $500,000 but over 20 years.

Ryssdal: Yeah.

Mullainathan: And so what people often do is say, "Hey, I don't want to wait 20 years. Give me all my money right now." So there are companies out there that say, "Fine we'll take your $500,000 spread out over 20 years and give you all your cash right now" at some ridiculous discount. That's a structured settlement. So there's a lot of money to be made.

Ryssdal: So $50 a click is nothing.

Mullainathan: No. Though ironically, you could go right now, do it, click away and you'd be charging these companies $50 a click.

Ryssdal: That's what I ought to do, right?

Mullainathan: Yeah, if you're ever bored.

Ryssdal: "You want to mess around with Adwords, pal? Here you go." Sendhil thanks a lot.

Mullainathan: Thanks Kai.

Ryssdal: Sendhil Mullainathan, he teaches behavioral economics at Harvard.

About the author

Sendhil Mullainathan is a professor of economics at Harvard University and Founder of ideas42, a nonprofit organization devoted to taking insights about people from behavioral economics and using it to create novel policies, interventions, and products.

Comments

I agree to American Public Media's Terms and Conditions.
With Generous Support From...