Reflections on Web measurement


[In 2002, the Tribune’s research director asked me to write about the brief history of understanding Web measurement for an NAA Research Federation publication. Herewith the result.]

By Owen Youngman, Chicago Tribune

“Once you have them by the eyeballs, their hearts and minds will follow.” Or so, not so long ago, it seemed.

Newspapers’ headlong rush onto the Internet in the late 1990s came at a real cost, one that is not enumerated merely by counting up dollars invested, spent, or dissipated. Even for large newspapers like the Chicago Tribune, the fact that we suddenly could be read by people all over the world obscured a number of interesting questions, including:

         Should we want to be?

         What I mean, of course, is that like many other companies that hit the World Wide Web in its first few years, newspapers found one of their rationales for being and staying online merely in the number of people they thought were in the audience. We congratulated ourselves for being sophisticated enough to move from “hits” to “page views” to “unique visitors” as we tracked our progress. 

Of course, we did that partly because it didn’t take very long to add up the short columns of advertising revenue—and partly because the sophistry involved in arguing that an ad that never scrolled into view should count as an ad “impression” was a little too much even for the strongest of stomach. 

There is an argument for worldwide availability; American newspapers’ fundamental role in a free society argues for broad dissemination of the work of our reporters and editors. But there is even a better argument for expanding our influence closer to home by understanding our intertwined and overlapping Web and print readers, and most of us didn’t remember to make it.

It’s bad enough that we didn’t take the time to build, test, and iterate systems for identifying who was in our local Web audience, perhaps because the true believers saw little benefit in testing nervous circulators’ hypotheses about “cannibalization.” In fact, when it came time to start measuring the local audience, we let vendors make inferences from sample sizes so small and so unscientifically selected that the results bore no more resemblance to reality than an Enron limited partnership agreement or a WorldCom balance sheet.

         But many of us also didn’t benchmark what our brands stood for offline so that we could measure how, and whether, being online might change that. We didn’t do much to figure out whether our journalism was having more impact, or less, when it became untethered from newsprint and ink. And we didn’t spend money when there was actually money to spend.

So today, we sometimes feel strategically adrift. When we try to assess the reach of the Chicago Tribune, chicagotribune.com, metromix.com, and chicagosports.com among 18- to- 34-year-olds in the Chicago market—whether in combination or in relationship to one another—I’m still pretty much just licking my finger and holding it to the wind. Do we have a chance to reach this audience in new ways? We’d know better if we could somehow know how many of the million people of that age who’ve read the Tribune this month are the same as, or different from, the 600,000 in that bracket who have used metromix.com.

Here’s what is missing: valid correlation. In Chicago, our ongoing project to understand the value that users and readers attach to our Web efforts goes by the name “Print and Internet Customer Relationships.” We want, we say, to “change what it means to be a Chicago Tribune subscriber.” Trouble is, we might not like our subscribers’ version of a new definition, because, in a sort of crypto-Heisenberg-Principle kind of way, while we weren’t measuring them they moved. Who knows; no benchmarks.

More and more Web sites are on the “required registration” bandwagon these days, I admit (and a brave few are testing the waters on requiring some sort of payment for access to content). Many that are not point out that we don’t know exactly who buys the newspaper at a newsstand, either. While conceding the point, we need not yield to it. 

Instead, we must ask the research community, and the experts in our midst, not just to correlate two kinds of readership, but also to link that correlation to action by our readers—whether purchase behavior, civic participation, or a decrease in the propensity to cancel. That is the data that strategists will need in the years just ahead to decide both how and how much to invest in efforts to keep us widely read in our backyards.

If then we’re also read widely in someone else’s backyard, we should all be able to live with it.

Owen Youngman, now vice president of development for the Chicago Tribune, was responsible for the creation and launch of the Tribune’s Web sites in the late 1990s.


About Owen Youngman

Professor Emeritus of Journalism and formerly Knight Chair in Digital Media Strategy, Medill School of Journalism, Northwestern University. Formerly senior vice president/strategy and development and director of interactive media, Chicago Tribune.