Stephen Baker on The Numerati, Google, Targeted Ads and the Greek God of Walks 2


Stephen Baker of BusinessWeek addresses a crowd at Google’s New York Offices in October

By Kyle Austin

RaceTalk recently had the opportunity to pull Stephen Baker, a senior writer with BusinessWeek, away from hawking his new book, to answer some questions on, well, his new book (among some other things). Here is a large excerpt of our conversation on the Numerati, Google, IBM, targeted ads, the credibility of the New York Times and even the “Greek God of Walks.”

RaceTalk: Thanks for joining us Stephen. I’ve done quite a few of these interviews with folks like John Markoff, Joe Nocera and David Kirkpatrick (to name a few), and you’re the first to schedule an interview via Twitter. It got me wondering, what was the experience like writing a story through Twitter earlier this year? I remember seeing the collaborative process online, but didn’t have a chance to catch the end result.

Stephen Baker: The story “Why Twitter Matters,” came out in May, a week after I opened up the live thread with the Twitter community, which allowed them to contribute to the story (May 8-9). Anyway, I got a lot of input, but the trouble with it was people didn’t want to write paragraphs. I was hoping they would craft the story with me. They just wanted to give me ideas and leave the story writing to me.

RT: I imagine it was tough for them to contribute 140 characters at a time?

SB: Yes. I thought they may try to write paragraphs together (140 characters at a time). When I set it up I thought it was going to be so easy. I would write the topic sentence for the paragraph and while the responses come in, I can go off to the Modern Museum of Arts and enjoy myself for half an hour – and then come back to see the paragraph constructed. It ended up being pretty chaotic and took a lot more work than a regular story.

RT: Seems like from the journalists I follow, you are updating more frequently on Twitter than most. Do you still enjoy it and find it useful in communicating with the people that you are most interested in keeping in touch with?

SB: I mean I enjoy it. I only update three or four times a day, so I don’t do it that much. But I do enjoy it, and it’s occasionally useful.

RT: Cool, so let’s get to your book on The Numerati.  Although, I haven’t gotten a chance to read the full book, I did read the forward and some of the very positive reviews. On Amazon.com it is currently ranked #2 in books on “Social Aspects” and #12 in “History and Philosophy.” In reading the forward, It sounds like Samer Takriti pushed you to launch research on the story, but how did you get interested in the numbers behind business (and even life) to begin with?

SB: I was not at all interested in starting this book project to begin with. The only reason I was looking at the subject was because I was looking into national competitiveness. Someone at BusinessWeek suggested that math was at the heart of the national competitiveness issue. In other words, we weren’t graduating enough mathematicians. Math was at the heart of all these technologies that the Chinese, Indians, Europeans and  Japanese were threatening our nation’s supremacy with. So the editor in chief of Businessweek sent me off to do a cover story on math with no further instruction. So I began looking at math and it is such an enormous world. I didn’t have a story on math until I spoke to Samer. Then I saw this data-mining and modeling, which was really important, not only to business, but society. It really gets at how we are understood; and we are understood by our data. Those people that can understand this data and master these tools have to do with computer science and mathematics (aka The Numerati). That’s how the project started.

RT: One of the first places you visited for the project was IBM’s Thomas J. Watson Research Laboratory, where they are studying worker’s data (i.e. time spent, productivity). From visiting there, what were your thoughts on their thinking around using data in this way?

SB: I think their thinking is extremely advanced. I think sometimes it’s misinterpreted to be heartless and big brotherly. However, their hopes are that they can understand their workers better by studying certain data. In doing so, they can put them in positions where they can thrive. Put them along side workers that they get along with and in situations that are more challenging and interesting. The premise of this, I don’t think is bad at all. I guess the part that I would be concerned about is that not all companies are as enlightened as IBM in this regard. So if they (IBM) produce tools that let companies monitor and model their workers, some companies will try to put us into a 19th century assembly line – even as information workers sitting at computers.

RT: Interesting, so that is kind of the future of where this may be going. If you look at where data-mining is today, and I thought it was a telling inscription on the cover of your book from Chris Anderson which illustrates where it is today, “A Must read for anyone who wants to understand life and business in the Google age,” what are your thoughts on online advertising and behavioral targeting? Obviously there has been a lot of commotion recently about behavioral targeting and the failure of NebuAd, Phorm and ISP tracking; so how has Google been so successful (in using data-mining and targeting) – while escaping the privacy debate?

SB: I think they’ve escaped it largely because people love to use Google. They use it and view it as a vital tool. They look at it as Google providing a service to them and less as providing valuable data to Google. That’s one thing. The companies that are going to thrive in this world are going to have to give us something (value) for our data. Google passes that test. The other test that Google passes is that they don’t really have to know us specifically. In fact, they can misunderstand us many, many times. If they give us 10 ads a year out of thousands that are of value to us, they make money on that. They don’t loose money by showing ads that don’t have meaning to us. That’s the real key for the Numerati. The companies that thrive are going to be those that don’t loose anything by getting it wrong occasionally.

RT: Do you think they have the upper-hand over competitors because it’s search based, with consumers posing questions that they want to see the answers/advertisements for? It seems that advertisers have trouble in targeting display ads online, because they don’t have consumers posing direct questions that they can answer/target with ads.

SB: If you look at what advertisers know about us – some of them look at the Web pages we look at and some look at the articles we click. All of those are indirect ways of figuring out what our interests are. In looking at these they can understand quite a bit about you and me, but trying to boil it down to exactly what we would want from that information is hard. Google doesn’t know that much about you or me from a search query. They know something specific, which is what is he/she looking for. That what, he or she is looking for is often something directly connected to the marketplace. A much tighter connection to money and more direct then the other ways.

RT: I read in one of the reviews of your book where you noted “blogs are real-time reaction,” versus putting out a poll or doing a focus group and then waiting for the answers to come in. This you said, provides a huge opportunity for the Numerati to mine blogs. Can you expand upon this?

SB: If you look at traditional marketing, they bring 12 people into a room for a focus group and pay them and the people monitoring the group a huge amount of money. They ask them a series of questions about a face cream, the soda they are drinking or the advertisement they are supposed to watch. The trouble in that is those people are supposed to represent the entire marketplace. What’s worse is those people’s answers may not be honest. They might be trying to please the monitors, answer politically correctly or be too embarrassed to admit what their real answer is. They don’t always provide accurate answers. The blogosphere is full of millions of people’s thoughts and feelings, expressed rather freely, about almost anything you can imagine. The Numerati are getting the tools to go through millions of blogs to figure out the approximate age and gender of the blogger. Then they can also analyze the sentiment of the blogger on certain subjects – whether it’s Barack Obama or a Coca-Cola ad, which allows them to come up with a back of the envelope trend on if certain ads/marketing has been successful with certain demographics.

RT: Living in Boston, I caught your recent slide-show contribution to Boston.com on 10 technology developments headed for the mainstream. I was specifically interested in your thoughts on Newscred, which tracks and analyzes the credibility of news. What are your thoughts around Newscred and the overall idea going mainstream?

SB: The idea has some power behind it. The public really cares about this issue but it is going to be hard to model it. People often feel that the magazines or publications which they object to aren’t credible. I don’t want to touch a red button issue here, but as a journalist if I wanted to affect the slant of a publication and get my point of view into stories – I think the most difficult place to do that would be the New York Times. I have lots of friends that work there and the editors there are incredibly rigorous. At the same time the New York Times is loathed by a large group of the population and considered a left-wing paper. Those people may say the New York Times isn’t credible, while another large group of us would say it is. There are other news publications that aren’t well known – say the Des Moines Register, which wouldn’t get as many negatives as the New York Times and could perhaps come out more credible than the Times. Whether it is or not, I don’t know.

RT: Another idea you mentioned going mainstream was Compulsion TV. I found that interesting, given that there was so much talk about this idea several years ago. Everyone wanted to talk about how in the future, women will be able to buy Jennifer Aniston’s sweater while watching an episode of Friends (may be a bit outdated fashion today). Do you think we are getting closer to this and who do you think in the TV realm will be the leader in the way that Google has been online in targeting ads?

SB: I don’t know who the leader is going to be. I think it’s going to be big. I know a lot of companies are digging deep into that. One of the people that I profiled in the book – Dave Morgan of Tacoda (sold to AOL) – is a leader in behavioral advertising and is looking next at targeted TV ads. I mean the day is coming, fairly soon, when you and your next door neighbor will be watching the same TV show, while getting different ads.

RT: Out of the people you spoke with for the book is there a top 5 Numerati? The rocket scientists of data-mining?

SB: It’s tough to rank them. The Google people are way, way at the top but I didn’t talk to them for this book.

RT: Is that because they declined?

SB: They talked to me when I was doing the BusinessWeek cover story in 2006. Then I told them “I’m writing this book, I’ll be at it for two years, can I have a bunch of your time?” They didn’t really give me a lot of time. So I spent more time at Yahoo!, which served the same purpose. I really wanted to talk with big Web companies that deal with oceans, and oceans, of data. I’m kind of glad I didn’t do Google because they are so widely covered and if I just came out with research I did there two years ago now, it would seem stale. As far as top 5 stars, I don’t really have a list.

RT: Finally, you’re a Phillies’ fan, do you have them over the Rays in the World Series?

SB: They’ve got a 40 or 50 percent chance.

RT: We’re mostly Red Sox fans here, so we’re still recovering.

SB: You know, I was cheering for the Red Sox in game 5 because they looked so bad. I thought maybe the Phillies have a better chance again them. Jason Varitek can’t hit anymore, Beckett has a problem with his oblique, Wakefield is hittable and Manny isn’t in the line-up. Then they come storming back and I’m thinking I don’t really want to deal with this club. Pedroia and Youkilis can just tire pitchers out. The Phillies aren’t deep in pitching. If Hamel started a game and Pedroia and Youkilis both have two ten-pitch at bats to start, it would kill the Phillies’ rotation. So I thought it would be better to have the match-up with the more free swinging Rays.

RT: You should have interviewed the Red Sox number’s guy for your book – his name is escaping me.

SB: Bill James. He did Money Ball already though. The Numerarti really focuses on how they understand us as people. There’s plenty of room for the Numerati to do other things in baseball, finance, medicine and all these other areas. I really focused on the people aspect. The voter, the shopper, the worker, the lover, the potential terrorist. Understanding us all as people.

RT: Great way to end, thanks for your time Stephen.

(BTW – Stephen even used the Numerati to do targeted online advertising for his book, so you may see one of these ads targeted at you on the Web after reading this post)


Leave a comment

Your email address will not be published. Required fields are marked *

2 thoughts on “Stephen Baker on The Numerati, Google, Targeted Ads and the Greek God of Walks