2020 reader survey results

Recently we invited y’all to complete a reader survey, to see how we’re doing and help us better understand the reasons for the year-on-year decline in our traffic that began 11 months ago. Read on if you’re interested in the results, and my tentative thoughts on what to do in response to them.

We got 316 responses, which tells you something already because that’s well down from the 396 responses we got to our last reader survey back in 2017. It’s not a random sample of our readers, obviously. But it’s probably a reasonably representative slice of our regular and regular-ish readers.

The respondents were geographically representative of our readership, because their distribution matches the distribution of our pageviews: 53% of respondents are based in the US, 10% Canada, 6% UK, 18% non-UK Europe, 4% Australia/New Zealand, 2% each from Brazil and India, the rest from elsewhere. All that’s nearly identical to our last survey.

Respondents were mostly a mix of grad students (26%, with the vast majority of those being PhD students), postdocs (19%), faculty for <6 years (14%), and faculty for 6+ years (21%). So our audience remains primarily academic, which is the audience we usually imagine when we sit down to write. Respondents also included 11% professional non-academic ecologists, and 10% other occupations. Compared to 2017, that’s a pretty substantial shift away from grad students and postdocs and towards faculty. Go back even further, and our regular readership used to be something like 40% grad students.

Respondents are mostly regular or semi-regular readers. 36% read most or all of our posts, and 49% read some of our posts; just 15% read few of our posts.

Respondents have mostly been reading us for years. 39% have been reading us for >2 but <5 years, 31% have been reading us for 5-7 years, and 8% started reading my blogging for Oikos Blog and followed me over here when I founded Dynamic Ecology ~7 years ago.  Just 9% started reading us within the last year, and 13% have been reading us for 1-2 years. Compared to our previous survey, that’s an appreciable shift in the “age distribution” of our regular readership towards longtime readers.

Longtime readers are the most regular readers. Almost all respondents who’ve been reading us for 5+ years (including those who started reading my work over at Oikos Blog) read at least some of our posts, and they’re the most likely to report reading most/all of our posts.

We aren’t picking up many new readers from among current postdocs and faculty lately. Almost every respondent who’s started reading us in the past year is a PhD student. None are faculty. That’s a change from the past, although we can’t say exactly how big a change. For instance, many of the respondents who’ve been profs for 6+ years must have started reading us as faculty or postdocs, given how long they’ve been reading us.

Few respondents think we’re getting better, or worse, but the overall opinion has deteriorated a bit since our last survey. When asked how the quality of this blog has changed over the time they’ve been reading it, most respondents gave some non-directional answer: 37% said it fluctuates, 28% said no change, and 25% said don’t know/not sure. Only 8% said we’ve gotten worse and 5% said we’ve gotten better. In our last survey in 2017, 12% said we’d gotten better and 3% said we’d gotten worse. As you’d expect, respondents who’ve been reading us for <2 years are the ones most likely to say that they’re not sure how our quality has changed over time. Other than that, there aren’t any strong associations between how long respondents have been reading us, and their opinion of how our quality has changed over time.

About half of respondents (47%) read us at least as often as they always have. That’s down from 63% in 2017. As you’d expect, respondents who think we’ve gotten worse over time are unlikely to report reading us at least as often as they always have (only 3/24). Conversely, 9/14 respondents who think we’ve gotten better report reading us at least as often as they always have. And as you’d expect, a large majority of respondents who’ve been reading us for less than a year read us at least as often as they always have.

The overall picture of the results reported up to this point is that our regular and regular-ish readership is in large part a cohort of people who started reading us many years ago, back when we were a hot new thing. 🙂 That cohort is slowly drifting away. Presumably, some have stopped reading us entirely and so didn’t respond to this poll. Though there is a small core of longtime readers who still read us avidly and perhaps always will. We’re still picking up some new readers, particularly among PhD students. But we’re not picking them up fast enough to replace loss of traffic due to longtime readers drifting away. (Aside: we’re also losing traffic from “one-off” visitors who arrive from search engines. Presumably we’re losing that traffic at least in part because Google’s PageRank algorithm downgrades old pages. Many of our posts on topics of broad and perennial interest are now 5+ years old. But this poll doesn’t speak to that source of traffic loss.)

So, if we wanted to do something about that, what could we do? Well, it’s not clear. Because respondents have many different reasons for reading us less than they used to. Respondents were invited to select all their reasons for not reading us as much as they used to from a long list of options, and also given a write-in option.

• A large majority of respondents who read us less than they used to gave multiple reasons why. There was no obvious association between different reasons for reading us less.
• 29% of respondents read us less than they used to because they’re busier than they used to be. Which means that more than half of the 53% of respondents who read us less than they used to, read us less than they used to because they’ve gotten busier. That makes sense, given that a substantial portion of the respondents started reading us at least a couple of years ago. During the intervening years, many of those respondents presumably moved up the academic career ladder, and so got busier. Or at least they feel like they’ve gotten busier, which comes to the same thing for purposes of people’s blog-reading choices.
• 15% of respondents (so, ~30% of those who read us less than they used to) read us less than they used to because Brian and Meghan post less often than they used to. To which, it’s not just many of our regular readers who’ve gotten busier, it’s also Brian and Meghan! (aside: I’m not that much busier than I used to be. In contrast to Brian and Meghan, I haven’t taken on any major new service or personal obligations since I started blogging.)
• 13% of respondents read us less because our topic mix has changed. On the other hand, 12% read us less because we’ve grown stale and are always posting on the same old stuff! 🙂 I look forward to those two equally-sized groups of respondents arguing in the comments about who’s right. 😉 Joking aside, obviously there’s no objective sense in which both groups can be completely right; either our topic mix has changed or it hasn’t! (And before anyone asks, no, there’s no obvious difference between those two groups of respondents in terms of how long they’ve been reading us.) I note this not to criticize anyone for being irrational, or failing to pay close enough attention to how we have or haven’t changed. Heck, if you asked me whether we’ve changed or stayed the same over the last 7 years, I’d say “yes”! That is, I think we’ve changed in some ways while remaining the same in other ways. Which to some readers feels like (undesirable) change, and to other readers feels like (undesirable) stasis. All of those subjective perceptions are equally valid! But they are subjective perceptions. It’s impossible for us to address all those valid but conflicting subjective perceptions in a way that would please everyone.
• 11% of respondents read us less because they increasingly read/use Twitter instead. Which confirms a suspicion of mine. At this point, I think everybody knows that social media has replaced many of the functions that blogging used to serve. And Twitter is designed to be engaging, attention-grabbing; reading blogs isn’t designed that way. It was always a question of when, not if, this blog would cease to defy the general trend and start losing readers to Twitter.
• Then there were a whole bunch of reasons for reading us less, each chosen by at most just 8% of respondents. I was a bit surprised that “it’s a pain to read blogs on my phone” wasn’t chosen by more people.
• Only 8% of respondents (so, about 16% of those who read us less than they used to) wrote in a reason for reading us less than they used to. No single one of those reasons was named by more than a small minority of those 8% of respondents, and most were given by only a single respondent. So the options offered in the poll covered most of the bases.
• Only a very few respondents said they’re not sure why they don’t read us as much as they used to.

Obviously, we’d prefer our traffic not to decline. Not because traffic is an end in itself (it’s not), or because traffic brings us tangible benefits like money or awards (it doesn’t), but because the more traffic we have the better we’re able to achieve some of our blogging goals. Have conversations on topics of interest to us, shape the direction of the field, give advice/suggestions to anyone who might want them, and so on. And obviously, in absolute terms we still get a lot of traffic (thanks everyone!) So it’s not as if we’re no longer achieving our blogging goals. We’re just not achieving them quite as well as we used to. Is there anything we can do about that?

I don’t know. Maybe not. I mean, there’s nothing we can do to make our readers less busy, or make our blog as engaging as social media.* There’s nothing Brian and Meghan can do to post more often. They already post as much as they can and wish they had time to post more. It’s hard for us to change our topic mix. We already post on all and only those topics we feel inspired to post on, and on which we think we have something to say that’s worth reading. Believe me, if I had a magic wand I could wave to conjure up inspiration to write on other topics, I would! Plus, there’s no single topic we could write about more (or less) that would change the reading habits of more than a modest percentage of poll respondents.

Perhaps one thing we could do is write posts about recently-published papers, or even preprints. We’ve only ever done this a few times (example 1, example 2, example 3, example 4). And we’ve only ever done it when there was something unusual about the paper that moved one of us to write about it. For instance, that second example discusses a pair of papers that arose from a formal debate at the ASN meeting. I’d previously blogged about the debate, and so I wanted to blog about the resulting papers too. And that fourth example is Brian discussing the “North America has lost 3 billion birds” paper that was all over the news recently. Discussing the current literature would certainly plug us in more to what many people are already paying attention to. Scientific discussion on social media is almost entirely about “news”, in the broad sense of “stuff that just happened”, a category that includes recently-published papers. Posting about recently-published papers would be a change from what we often do, which is post on stuff that is interesting or important despite the fact that it’s not “news”. But I don’t think posting about recent papers would mean ceasing to talk about the sort of stuff we usually talk about. Because we’d be able to connect the papers we were writing about to broader issues we’ve written about in the past, longer-term trends in the field, etc. Note that I definitely don’t see us just writing paper summaries; that’s what journal-associated blogs are for (not to mention abstracts!) And I definitely don’t see us doing technical critiques of individual papers, which is what I think of when I think of “post-publication review”.

It’s worth noting that the few remaining widely-read scholarly blogs with stable and really large readerships are mostly about “news” in the broad sense of “stuff that happened recently”. Think Marginal Revolution, or Andrew Gelman’s blog.

The easy thing about that sort of blogging is that one never lacks for post topics. The hard thing about that sort of blogging is that it has to be timely, and it’s hard to be timely when you’re also trying to be substantive and interesting. Though I’m unsure exactly how timely or substantive it would have to be. For instance, Andrew Gelman often posts about stuff that happened months ago, and often comments only very briefly. Guess there’s only one way to find out! So watch this space.

The other thing I think we might try is to write more about topics that are specifically of interest to faculty, including post-tenure faculty. We’ve only ever done a bit of that. It seems like there’s not much public discussion about the experiences of post-tenure faculty, or online advice aimed specifically at them. But speaking as someone who’s been a full prof for several years now, I don’t feel like I need less advice than I did as a grad student or postdoc or pre-tenure prof. I just need advice on different stuff. So maybe we’ll start writing some!

Looking forward to your comments.

Image source.

*Well, I guess we could write deliberate clickbait, but that would amount to treating traffic as an end in itself. Which would be pointless.