Let’s not be so quick to trash the website statistics dashboard

(Image by Mad cherub via wikimedia)

Brace yourselves, my fellow journalists. Today we’re going to talk about … numbers.

Start with Stijn Debrouwere, who says:

Pageviews is a vanity metric: something that looks really important but that we can’t act on and that tell us nothing about how well we’re actually doing.

This is from a speech that Mathew Ingram later highlighted, largely with approval, on GigaOm. Debrouwere’s larger point is that online news media can do stupid things with the data they get about traffic, and may not understand what’s important. I agree with that, but it’s hardly news. Editors have done stupid things long before there was an internet, particularly with numbers. I’ve written before about my disdain for the typical newspaper readership survey.

Eventually, Debrouwere gets around to offering solutions, and they’re good ones. But before I get to those, let’s look more closely at the statement above, which Ingram liked enough to turn into a tweetable.

Pageviews “tell us nothing about how we’re actually doing?” In the aggregate for a site, that statement has some merit. Using total pageviews for a site to make decisions about posts is as much like digging through goat entrails as using print circulation.  That’s why, when I took over as Online Editor at The Plain Dealer, I quickly ended my predecessor’s habit of announcing at each morning’s news meeting the total views and unique users for the preceding day. Instead, I focused on individual stories.

At that level, pageviews provide a measure that tells us a lot about how we’re doing. It was from exploring pageviews at the story level that I saw our food coverage did poorly overall, but stories about restaurants opening or closing were strong. Or that we didn’t have much of an audience for TV reviews, but anything about local TV anchors took off.

Can we act on that? Absolutely. And not just by writing more of those stories. We can make sure that we don’t bury news of restaurant openings or closings in the middle of long notes columns, but rather give them their own headlines, for example.

And what I said up above about total pageviews? It’s true in general. But even there, you can spot obvious trends. In fact, the example Debrouwere cites proves the opposite of what he says:

Iowans love local news. They love it because they get the latest forecasts, traffic reports, a heads-up on school closings and a myriad of other information that makes your life so much easier when the world around you has frozen over.

When the weather is particularly bad, pageviews are particularly high.

But then March becomes April and April turns to May, and pageviews go down, and then go down some more.

But as of yet it’s impossible to manipulate the weather, so seeing those pageviews go down is not something you can really do anything about. You can set all the performance targets you want, but you can’t well fire your editor for not turning rain into snow.

No, but you can fire him if he fails to see that the numbers suggest he should pour on the weather news when it’s snowing, live-blog the commutes, and in myriad other ways exploit that reader interest when it surges.

Pageviews by themselves are a rough measure. As Debrouwere wrote earlier, they may be deceptive as a tool for long-term strategy. But the industry has been working with rough, deceptive measures for a long time; why should online be held to a higher standard? And pageviews offer a measure that’s easy for print journalists to understand: This is how many people at least glanced at your story. Sure, you’ll get reporters who say their stories are too important to be judged on pageviews. But if you repeat this to them often enough, it often sinks in: If a story is published but hardly anyone reads it, it can’t make an impact.

As I’ve said before, if you’re not already using traffic numbers, including pageviews, to figure out what readers really care about, you’re a fool. I will agree with Debrouwere, though, that you can use those numbers foolishly. Again, that’s hardly news. Statistics can always be abused — remember Mark Twain’s “lies, damn lies and statistics.” Particularly with pageviews, surface impressions must be explored more deeply.

Sports stories are commonly strong, especially in cities with pro teams. In Cleveland, the Browns dominated our pageviews in a way they haven’t done on the field since the early ’60s. We could post just about anything about the team and it would be big numbers — and so, yes, we posted just about anything. But I (reasonably politely) disagreed with another consequence of those stats: A desire on the part of our website folks to plaster the homepage with Browns headlines.

Despite what you may have heard about search engines and social media, the homepage was — at least for us — the single biggest driver of story traffic in general. However, that wasn’t nearly as true for sports stories, and especially Browns. We got a lot of those views off the main sports and team index pages. Sports fans knew where to find the news they were most interested in. I argued (politely, I think, but largely unsuccessfully) that by making our homepage so heavily Browns, we were doing little to increase traffic overall, and we were cheating other stories that would be invisible to most users once they fell off the homepage.

Debrouwere seems to be saying that looking at an overall dashboard of stats will lead to mistaken ideas, and keep journalists from doing the deeper digging to fine-tune their decisions. I say that looking at a dashboard gives you an overall perspective that can inspire further digging.

Ingram offers another note:

Media companies aren’t trying to bring back something they already had by using analytics — it’s more like they were remote villagers hidden in the rain forest who had never seen a ruler or a scale for measuring weight before, and suddenly when the web came along they were handed these tools and didn’t really know what to do with them.

This is a good point. But rather than an argument against using dashboards, I say it’s an argument for using them responsibly.

For more than a century, as journalists have gotten (I would argue) less and less like the majority of their readers, they have had no way of seeing the impact of that. For decades, circulation grew mostly on the strength of population growth and a shrinking number of papers to choose from. Even when the numbers started to drop, we could blame that on demographics and technology. All the while, we could maintain the happy fiction that every story in each day’s paper was read by everyone who bought up a copy — and, according to the “readership” numbers that the industry trots out whenever it gets too depressed to talk about circulation, by two or three times that many from pass-alongs.

Along came the internet, and the harsh truth emerges. Only 200 people read my story? Really? Yes, really. And while print and online audiences are different in significant ways, I’m betting that the story from a big-city paper that gets 200 (or fewer) pageviews online wasn’t getting a lot of interest in print, either.

Looking at a Google Analytics or Omniture dashboard at least has the value of waking print journalists up to reality. Looking at pageviews for individual stories can start conversations about why some stories do well while others don’t, and what can be done about it.

Debrouwere has more to say about this, and so do I. That will come tomorrow.