Yesterday I had an interesting Twitter conversation with Alexis Madrigal, now Technology Editor at The Atlantic, after he posted the following:
@loisbeckett @mat for us, there is a positive correlation between word counts and pageviews. I.e. More words = more views
I asked him if he could say more, and he did:
@jonathanstray I obviously can’t share too much, but at @TheAtlanticTech, we’ve seen our longer pieces — on average — do better.
@jonathanstray What I can’t say is whether that’s specific to The A. (Though my experience with science stories at Wired was similar.)
@jonathanstray And I should note also that my take is a gestalt impression. We haven’t done rigorous word count/traffic statistical work.
Madrigal also notes that The Atlantic doesn’t paginate the long stories, so he’s talking about actual story view numbers, not raw page views. Slate has also reported good traffic for their long stories. This has me wondering: where did the idea that short stories do better online come from? Who has data or experience to back this up, or is this common knowledge actually an urban legend?
I’d be willing to share relative data for WSJ.com, but the problem is forming the dataset. Because pageviews and word counts aren’t in the same place, we need a small enough dataset to keep this manageable. But it also has to be large enough to be significant and random enough to be valid. (I don’t have a good way to randomly select content from among everything we publish, though I do have a WSJ firehose.) If you could solve these issues and return me URLs with word counts, I could give you relative traffic for each. We might not be able to expose the whole dataset, but I think the results would OK to share with all. And, actually, I have no idea what we’d find…
Fascinating. And it’d be relatively simple to pull data from either side. I look forward to seeing what you come up with.
PS You should install Subscribe to Comments so we can get email notifications of new comments 🙂
Also, beyond page views, I’m curious how story length correlates to time spent on an article page.
Large factor is probably because shorter stories are often duplicate content that people have already read and skip. While longer stories are often original content.
由于我完全不同观测，一些专业词汇就看的似懂非懂的。所以想请教一下下面这句话的正确的翻译：”Mass loss from cool giants and suapngierts is evidenced by P Cygni profiles in optical spectra, by the 10 μm silicate feature in IR spectra, and by maser line emission from OH and other molecules in radio-frequency spectra.“