We (and all other websites) are not nearly as widely read as you’d think from our traffic stats (CORRECTED)

Just a quick navel-gazing post that will be boringly familiar to many of you but might be eye-opening to others.

As you probably know, this blog gets a lot of traffic for a science community blog: about 60,000 pageviews and 25,000-30,000 unique visitors per month. (Thanks for reading, everybody!) But as some of you may not know, those numbers give a very overinflated impression of our audience size. Because guess how long the average visit to Dynamic Ecology lasts? Seriously, guess!

No, lower. πŸ™‚

if-you-could-0jshhc

Try approximately 60 seconds. Which isn’t long enough to read even one of our typical posts.

Ok, those data come from a third party visitor-tracking website rather than WordPress, so they’re probably not spot on. But I know they’re close, because the same is true for pretty much every website. For instance, the average visit to Andrew Gelman’s blog lasts less than 2 minutes; the same is true for the average visit to FiveThirtyEight and Vox. Heck, it only goes up to about 3 minutes at the New York Times, which is mostly paywalled and gets much of its traffic from subscribers who you’d think would visit for extended periods.

In fact, it’s even worse than the average visit length makes it look. Because a large fraction of visitors to Dynamic Ecology, or any website, only stay for a few seconds. Whether because they found us via search only to discover we weren’t what they were looking for, or decided that that day’s post title didn’t sound interesting, or whatever. The average visit length is longer than the modal length (while remaining short in an absolute sense) because only a fraction of visitors stay long enough to read something.

So if you want to know how widely read this (or any) website is, the answer is “Not nearly as widely read as you’d think, just looking at the traffic data.”

Similar skewness shows up in our unique visitor data. I said above that we get about 25,000-30,000 unique visitors per month. But we typically get about 8,000 per week, which only adds up to our monthly unique visitor count if the large majority of visitors in any given week are making their only visit of the month. And we get about 350,000 unique visitors per year, which implies that most of the visitors who visit us in a given month are making their only visit of the year. (Correction: WordPress determines the yearly unique visitor total by summing the monthly totals. That doesn’t affect the main point of this post, though.) There probably aren’t more than a few hundred people,Β maybe 1000 at most, who read us regularly (for any reasonable definition of “regularly”). Again, I’m sure something similar is true for almost any website.

Which is fine, by the way; I’m not bemoaning this state of affairs! It is what it is, and there’s no changing it unless you’re Facebook. I remain flattered, humbled, and proud that we’re as widely-read as we are. But I find it a useful reality check to remind myself that only a small fraction of ecologists (or scientists, or academics, or etc.) read us more than once in a blue moon.

25 thoughts on “We (and all other websites) are not nearly as widely read as you’d think from our traffic stats (CORRECTED)

  1. “There probably aren’t more than a few hundred people, maybe 1000 at most, who read us regularly (for any reasonable definition of ‘regularly’).”

    I think there might be an easy way (*cough* poll *cough) to see how true this might be. A simple one-question poll could help elucidate exactly how regular your readership is (apologies if you’ve done this and I’ve just forgotten about it!)

    • We have polled readers before on this (along with other questions). We never get more than 400 respondents. And most of those people say they read “some” or “most” of our posts. So that’s one estimate of our regular readership–400ish people.

  2. I wonder how they track average visit length. Depending on the methodology, my most of my page views would either end up looking like a bunch of really short visits, or one outrageously long one.

    My reading mode for this, and most other “blog-like” web content is to open new tabs for pages that look interesting as I hit things, and I rarely spend the time to actually read the pages when they load. I revisit my collection of open tabs and read in snippets when I’m stuck waiting for some calculation to run, or when I’m contemplating my coffee/etc.

    As a result, I’ve got tabs for some of your posts that I’ve yet to finish, that have been open for months, but I haven’t spent more than a couple minutes at a time with any of those tabs in the foreground.

      • Heh – I still haven’t finished “Favorite Novels Featuring Scientists…” I’ve got a handful of other posts here currently open too, but I can’t guarantee those tabs all date from the original date of posting.

    • Depending on who reports this, there are a few standard ways to determine if someone is idling on a page, or actively browsing a site. Sites with lots of JavaScript or other active elements (e.g. Facebook) can actually keep track of how long you are actively using that tab. Static webpages typically compute average session duration based on if a user’s IP queries several pages on the site within a reasonable amount of time. If you sit on a page for 30 minutes, many of these will time out so that if you click on a link, it will be a “new session.” Here is how Google computes this, for instance:
      https://support.google.com/analytics/answer/1006253?hl=en

  3. 400 regular visitors is a nice audience. 1000 is amazing. What is a unique visitor? For example, I visit this site from my laptop, work desktop, phone, and ipad. And occasionally a computer in a classroom. Am I four (or more) unique visitors?

    • Yes, unique visitors are IP addresses (I think). So yes, you’re four+ unique visitors. That’s another reason why our unique visitor count overestimates how many people read us regularly.

      • This also works the other way too, I think. Since IP addresses are determined by wireless networks, I think two computers at the same university using the same wireless network will be treated as one visitor unique visitor. Maybe these two complications cancel each other out πŸ˜‰

      • “I think two computers at the same university using the same wireless network will be treated as one visitor unique visitor. ”

        Really?! I had no idea. Wow, if that’s right it would make the unique visitor data pretty much uninterpretable.

  4. Do you know the traffic statistics for older pages? I keep revisitng them to retrieve information every now and then…and the duration of my stay is much longer than a few minutes

    • Yes, WordPress tells us the number of visits to each post each day. Our most popular posts all time are those that have been getting a steady stream of 20-80 daily visits for years, because they come up on the first page of google search results on searches related to the post topic. They’re mostly advice posts. A somewhat outdated list: https://dynamicecology.wordpress.com/2016/10/27/our-greatest-hits/

      But on any given day a decent fraction of our old posts will get at least one visit.

  5. I frequently visit just to check whether there is any new post, and close the tab pretty fast if there is none. Not sure how common this kind of behavior is, but I guess it reduces drastically both mean and mode.

  6. There is a rule-of-thumb for interpretive exhibits in visitor centers and museums–3 seconds to attract their attention with a headline, 30 seconds to convey the message, and 3 minutes for those really interested in the details. That’s if you are lucky; then they are ready to move on. And this was long before the internet.

  7. It seems that a better measure may be how many followers the blog has, and how many comments you get, perhaps. I am a follower, and I often comment or at least like if I think the post is interesting to me.(That also means I spent more than 60 seconds reading it!) I’d estimate that I probably read fully one of 5-10 of your posts, depending on the month. That is not a comment on the posts I don’t read, but simply that they are not relevant or of interest to me.Fourhundred regulars is just something that I could dream of for my blog, but then I don’t write mostly about personal experiences or naturalist stuff, so I can’t expect that anyone other than friends would be even remotely interested. For what it’s worth, I very much like this blog as it is generally very stimulating.

    • Thanks for the kind words. I’m sure your approach–only reading the occasional posts that fall within your areas of greatest interest–is common. It’s certainly my approach.

      Re: our follower count as a measure of our audience size, that’s difficult to interpret, unfortunately, for a few reasons. First, *many* of the followers listed in our WordPress “follower” count are bots. Second, the WordPress “follower” count attempts to include Twitter followers, but last time I checked (a long time ago) it wasn’t counting them accurately; it was off by thousands! Third, our list of email followers likely includes lots of subscriptions to now-defunct email addresses. Fourth, lots of people and organizations that follow our Twitter feed (which is mostly just a robot that announces new posts) never interact with our tweets. Never “like” them, never retweet them, never reply to or subtweet them, and never click the link to the day’s post. Basically, as best I can tell, lots of Twitter users casually decide to follow our Twitter feed but don’t really get anything out of doing so and never bother to cull us from their follows.

  8. Pingback: El lanzamiento

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.