Online Journalism: Editing by the numbers

Andrew Grant Adamson pointed me towards Lucas Grindley’s post on trying to come up with a formula to decide which stories get web coverage

A growing culture change and fear that the printed page might die are sending more reporters and editors to my desk. Everyone has an idea; a lot of which are good. Here’s how I decide which projects make the list, and in what order.

His method? Grindley’s page impression formula

First, estimate how many page views the project will earn per quarter. Let’s say Project X is estimated to earn 1,000 page views per day, which is 91,000 page views per quarter. Then estimate how many hours of work it will take for the team to complete the project. Let’s say Project X requires 40 hours of your team’s time to create. Now divide to find the rate of page views per work hour.Project X generates 2,275 page views per work hour. (91,000 / 40 = 2,275)

Lucas is careful to say that ‘good journalism’ should transcend the formula but this whole idea of guessing, and it is guessing isn’t it, the popularity of a piece seems unnecessarily, well, numeric.

The numbers of page impression/view is only really of interest to the advertising who can charge per serving. It might give you an indication of popularity but it is less than reliable, especially during working hours.Without a lot more external information the page impressions are misleading at a site centric level.

At what’s your benchmark? What constitutes a popular piece and do you only take editorial ideas on that basis.

Take this example from Matt Waite from the St. Petersburg Times:

This morning, my employer ran a story about a poor girl who has had the hiccups for the last three weeks (look for her on the Today show Friday morning). Readers have responded in amazing and massive ways. Mary Jane Park, the reporter who wrote the story, was stunned this morning when she came in and had 148 voice mail messages waiting for her. People by the hundreds — nearly 300 by mid afternoon — wanted to call her and give cures for the hiccups.

No page impressions there No mention of page impressions there, but a really popular story. Does that mean only fluffy hiccup stories make the grade.

It’s a little unfair to question the formula, I know. I think this is just as likely to work because of Lucas’ experience of what works for his audience as it is the numbers.

But, bearing in mind Rob Curley’s comments on the pressure to monetise content, I would be very surprised if anyone would stand for this kind of formulaic editorial decision making in print.

Just because we can get the data, should we be such slaves to the numbers online?

4 thoughts on “Online Journalism: Editing by the numbers”

  1. Whoa, whoa, whoa.

    Who said the hiccup story didn’t get page views? I disagree. Let’s ask Matt to volunteer the page view per work hour for that story and see what you find. My guess is it’s pretty good compared to other multimedia projects like this one.

    Page views are not misleading as an indicator of your audience’s interest. That is, unless you’re using something like Ajax that doesn’t rely on page views. Still, even in that instance, compare apples to apples and you’ll be OK. If you’re tryin to decide which Ajax application to develop, you can compare page views per work hour to find out which one you think will be most worth doing.

  2. Fair point and perhaps I do need to clarify a little. I meant no page impressions mentioned.

    In the quote, there was no information about page views, there was a different measure of popularity – phone calls. You’re right. There may have been a lot of page views as well. My point, in raising the story, was that given its poularity, however it was measured, was it a marker for the general tone of content.

    What drives what? If a certain type of story drives the page views up will your team to only turn up at your desk with fluffy stories?

    I accept that page impressions, relative as they can be, are an indicator of popularity, even with their weaknesses(especially singular site centric ones). But I’m aware that I’m deliving deep in to an element of, not the whole point of the post.

    Any way we can use to make decisons on what is going to work online is great to add to the mix and I know you arent advocating a sole reliance on figures. My broad point was that numbers, especially estimates, shouldnt be the only or overriding one.

  3. I don’t know what the pageview counts are, but the work took me about an hour and a half. Half an hour was spent getting someone to show me how to connect the phone, the mixing board and the Marantz recorder together, which I had never used before. The actual editing took less than 15 minutes. And here’s the sum total of training I’ve been given on SoundStudio, the software I used: IT guy: “Hey, I installed SoundStudio on your Mac.” Me: “Cool.” So, really, it could have gone much faster. For the sake of easy math, well call it one hour of work against a bajillion page views, which I know it’s getting, judging by the flood of comments, emails and phone calls we’re still getting. So, in terms of time invested, the unrelated story meeting I had to go to before I could tackle the audio lasted longer and got zero page views (thank God).

Leave a Reply