A recent article in The Times about the flawed way that experts predicted health care costs had an extra element when read online: personalized location.
A few paragraphs down, readers hit a sentence that read something like this: “Consider New York City, our best guess for where you might be reading this article.” (An accurate guess for me, although not one requiring rocket science — or even data science.)
Information about the reader’s likely location followed. A bit farther down, the article, which also ran on the front page of Tuesday’s paper, offered an option to insert another location into an accompanying map. I could see, for example, how costs compared in my hometown, Buffalo.
It was an impressive feat of data visualization.
David Leonhardt, editor of The Upshot, told me he sees it as an example of how much news presentation has improved. Praising the work of the Upshot staffers Kevin Quealy and Margot
Sanger-Katz, he wrote to me:
“I find it amazing to think about how much better this story is today than it would have been even 10 or 15 years ago, when the graphics probably would have been a black-and-a-white national map in the paper and maybe a chart or two.”
Others shared his view. Mike Kelly of Austin, Texas, wrote: “The graphics included in the article online comparing the readers locale are useful and sort of breathtaking in the application of reader data. By reporting on the price of care as if there were actually a market in care, you create understanding among your readers that otherwise wouldn’t be available.” And on Twitter, Sarah Gollust praised the piece.
But some readers told me that they found it off-putting. Robert McGrath wrote:
Midway into the article, I was shocked to read: “Consider Urbana, Ill., our best guess for where you might be reading this article.” The following text also included data plots that include Urbana, IL.
This is a pretty good guess, given that it is where I receive the paper edition. And I see how the sort of clever algorithm generated a “customized” version of this article, “just for me”. The question is, “why?” I did not expect or request this silly “customization”. It’s a really creepy experience, frankly.
Mr. McGrath wondered whether he had missed the chance to opt out of this, and said he had no interest in personalized articles or graphics.
Another reader, Alec MacLeod, had a similar reaction. “What does this mean for the integrity of stories and consistency for all readers?” he asked. And he added, simply: “I don’t like it.”
Mr. Leonhardt told me that the localization was achieved through using the reader’s IP address, and that it was not much different from the kind of geographic zoning newspapers have done for many years. Subscribers who live in a city’s northern suburbs receive a different edition of the paper than those who live in the southern suburbs. In fact, the health-care article probably offers less specificity than newspapers’ county-based zoning, he said.
It’s not the first time that The Upshot has used reader location to broaden the scope of its offerings. Another example, from last spring, was an interactive map on the best and worst places to grow up.
What about the readers who find it creepy? “I’m sympathetic to the notion that you can go too far with Big Brother technology,” Mr. Leonhardt said. “I’m not a fan of going to extremes on this.”
But, he said, the approach used in this article and informational graphic isn’t an example of that. (A side note: Mr. Leonhardt soon will leave The Upshot, which he has led since its beginning last year; he will become a staff columnist on the Op-Ed pages of The Times. His optimism about the news business in the digital age is reflected in an informative presentation he gave at Google earlier this year.)
For both journalistic and business reasons, The Times is experimenting with various kinds of personalized news. Online readers can expect to see more of it. (One small example, right now, is the “Recommended for You” listing of articles on the home page, near the “Most Emailed” and “Most Viewed” lists.)
In some ways, it’s risky. A New York Times that gives readers too much personalization could become more of an echo chamber than a news provider. (Eli Pariser, author of “The Filter Bubble,” lays out the dangers to democracy if citizens’ worldviews are narrowed by excessive personalization of news and information.) My sense, after a number of conversations, is that Times editors and product experts are well aware of the potential downsides, and are striving to avoid them.
Although it’s a fact of life that using the Internet means losing a certain amount of privacy, I understand why some readers were put off in this case. The Times could have quite easily provided readers with an opt-in: “Want to see results for your area? Click here.”
As the paper continues down this path, it’s important to do so with awareness and caution. For one thing, some readers won’t like any personalization and will regard it as intrusive. For another, personalization could deprive readers of a shared, and expertly curated, news experience, which is what many come to The Times for. Losing that would be a big mistake.