One of our yearly traditions here is to thank all y’all CSS-Tricks readers at the passing of a new year. It means a lot to me that people come here and read the words I write, and the words of all our staff and guest authors that contribute here as well.
Plus, we dig into the numbers this time of year. I’ve always tried to be open about the analytics on this site. Looking at them year after year always serves up a good reminder: niche blogging is a slow game. There’s no hockey-stick growth around here. Never has been, never will be. The trick is to build slowly over time, taking care of the site, investing in it, working hard, and with some luck, numbers trend upward. This year, overall traffic didn’t even do that. Sometimes you gotta fight for what you’ve got! Growth came in other areas though. Let’s take a gander.
It was January 1st, 2019 that the current design of this site (v17) debuted, so this entire year overlaps perfectly with that. I’ll certainly be tempted to release major iterations with that same timing in the future for comparison sake.
Google Analytics is showing me 90.3 million pageviews, which is a bit of a decline from 2018 at over 91 million. A 1% decline. Not a big problem, but of course I’d way rather see a 1% increase instead. We’ll take that as a kick in the butt to produce a stronger year of content to hopefully more than win it back.
Looks like we published 726 articles over the year, which includes link posts and sponsored links. A good leap from 636 last year and 595 the year before that. Clearly quantity isn’t the trick to traffic for us.
I don’t know that we’ll slow down necessarily. I like the fact that we’re publishing multiple times a day with noteworthy links because I like to think of us as a timely industry publication that you can read like a daily or weekly newspaper in addition to being an evergreen reference. I don’t think we’ll invest in increasing volume, though. Quality moves the needle far more than quantity for this gang.
There is a bunch of numbers I just don’t feel like looking at this year. We’ve traditionally done stuff like what countries people are from, what browsers they use (Chrome-dominant), mobile usage (weirdly low), and things like that. This year, I just don’t care. This is a website. It’s for everyone in the world that cares to read it, in whatever country they are in and whatever browser they want to. We still track those numbers (because Google Analytics automatically does), so we can visit them again in the future and look historically if it gets interesting again. Taking a quick peak, however, it’s not much different than any other year.
Performance numbers are always fascinating. Google Analytics tells me the average page load time is 5.32s. On my fast home internet (even faster at the office), the homepage loads for me in 970ms, but it’s more like 30 seconds when throttled to “Slow 3G.” “Fast 3G” is 8 seconds. Sorta makes sense that most visitors are on faster-than-3G connections since the traffic is largely skewed toward desktop. No cache, we’re talking 54 requests (including ads) and 770KB (fits on a floppy). It’s good enough that I’m not itching to dig into a performance sprint.
You’d think we would do a section like this ever year, but because of our URL structure, I haven’t had easy access to figure this out. Fortunately, in March 2019, Jacob Worsøe helped us add some Custom Dimensions to our Google Analytics so we can track things like author and year with each pageview.
That means we can find things, like the most popular articles written in 2019, rather than just the most popular articles looked at in 2019 — regardless of when they were was written. Here’s a graph Jacob sent:
Here’s that list in text:
- The Great Divide
- Change Color of SVG on Hover
- An Introduction to Web Components
- Where Do You Learn HTML & CSS in 2019?
- The Many Ways to Change an SVG Fill on Hover (and When to Use Them)
- Look Ma, No Media Queries! Responsive Layouts Using CSS Grid
- How to Section Your HTML
- Prevent Page Scrolling When a Modal is Open
- CSS Animation Libraries
8.25% of traffic came from articles written this year. If you look at where these articles fall on the list of all URLs in 2019 (not just those published in 2019), the top article starts at #75! Hard to compete with older articles that have had time to gather SEO steam. This kind of thing makes me want to get re-focused on referential content even more.
Interesting that our top article was editorial, but everything else is referential. I like a little editorial here and there, but clearly our bread and butter is how-to technical stuff.
There are two aspects of search that are interesting to me:
- What do people search for right here on the site itself?
- What search terms do people use on Google to find this site?
On-site search is handled by Jetpack’s Elasticsearch feature, which I’m still quite liking (they are a sponsor, but it’s very true). This also means we can track its usage pretty easily using the analytics on my WordPress.com dashboard. I also installed a Search Meter plugin to track search form entries. I can look at Google searches through the SiteKit plugin, which pulls from Google Search Console.
Here are all three, with duplicates removed.
|Jetpack Search Data||Search Meter Search Data||Google Search Data|
|3||css tricks||flex||css tricks|
|4||flexbox guide||animation||css important|
|5||css grid||svg||css triangle|
|6||css flex||position||mailto link|
|7||grid guide||css grid||vertical align css|
|8||css important||css||css comment|
|9||the great divide||border||css shapes|
|10||css shapes||background||css background image opacity|
There is a bit of a fat head of traffic here with our top 10 pages doing about 10% of traffic, which syncs up with those big searches for stuff like flexbox and grid and people landing on our great guides. If you look at our top 100 pages, that goes out to about 38% of traffic, and articles past that are about 0.1% of traffic and go down from there. So I’d say our long tail is our most valuable asset. That mass of articles, videos, snippets, threads, etc. that make up 62% of all traffic.
It’s always this time of year I realize how little social media does for our traffic and feel stupid for spending so much time on it. We pretty much only do Twitter and it accounts for 1% of the traffic to this site. We still have a Facebook page but it’s largely neglected except for auto-posting our own article links to it. I find value in Twitter, through listening in on industry conversations and having fun, but I’m going to make a concerted effort to spend less time and energy on our outgoing social media work. If something is worth tweeting for us, it should be worth blogging; and if we blog it, it can be auto-tweeted.
But by way of numbers, we went from 380k followers on @css to 430k. Solid growth there, but the rate of growth is the same every year, to the point it’s weirdly consistent.
I also picked up an Instagram account this year. Haven’t done much there, but I still like it. For us, I think each post on Instagram can represent this little opportunity to clearly explain an idea, which could ultimately turn into a nice referential book or the like someday. A paultry 1,389 followers there.
I quite like our newsletter. It’s this unique piece of writing that goes out each week and gives us a chance to say what we wanna say. It’s often a conglomeration of things we’ve posted to the site, so it’s an opportunity to stay caught up with the site, but even those internal links are posted with new commentary. Plus, we link out to other things that we may not mention on the site. And best of all, it typically has some fresh editorial that’s unique to the newsletter. The bulk of it is done by Robin, but we all chip in.
All that to say: I think it’s got a lot of potential and we’re definitely going to keep at it.
We had the biggest leap in subscribership ever this year, starting the year at 40k subscribers and ending at 65k. That’s 2.5× the biggest leap in year-over-year subscribers so far. I’d like to think that it’s because it’s a good newsletter, but also because it’s integrated into the site much better this year than it ever has been.
Oh, bittersweet comments. The bad news is that I feel like they get a little worse every year. There is more spam. People get a little nastier. I’m always teetering on the edge of just shutting them off. But then someone posts something really nice or really helpful and I’m reminded that we’re a community of developers and I love them again.
4,710 approved comments. Up quite a bit from 3,788 last year, but still down from 5,040 in 2017. Note that these are approved comments, and it’s notable that this entire year we’ve been on a system of hand-approving all comments before they go out. Last year, I estimated about half of comments make it through that, and this year I’d estimate it at more like 30-40%. So, the straight-up number of comments isn’t particularly interesting as it’s subject to our attitude on approval. Next year, I plan to have us be more strict than we’ve ever been on only approving very high-quality comments.
I’m still waiting for WordPress to swoon me with a recommitment to making commenting good again. ;)
There were a couple of weeks just in December where I literally shut down the forums. They’ve been teetering on end-of-life for years. The problem is that I don’t have time to tend to them myself, nor do I think it’s worth paying someone to do so, at least not now. Brass tacks, they don’t have any business value and I don’t extract enough other value out of them to rationalize spending time on them.
If they just sat there and were happy little forums, I’d just leave them alone, but the problem is spam. It was mostly spam toward the end, which is incredibly tedious to clean up and requires extra human work.
I’ve kicked them back on for now because I was informed about a spam-blocking plugin that apparently can do incredible work specifically for bbPress spam. Worth a shot!
Interestingly, over the year, the forums generated 7m pageviews, which is 7.6% of all traffic to the site. Sorta makes sense as they are the bulk of the site URLs and they are user-generated threads. Long tail.
✅ Polish this new design. Mixed feelings. But I moved the site to a private GitHub repo half-way through the year, and there have been 195 commits since then, so obviously work is getting done. I’ll be leaving this design up all of 2020 and I’d like to make a more concerted effort at polish.
✅ Improve newsletter publishing and display. Nailed this one. In March, we moved authoring right here on the site using the new Gutenberg editor in WordPress. That means it’s easier to write while being much easier to display nicely on this site. Feels great.
☯️ Raise the bar on quality. I’m not marking it as a goal entirely met because I’m not sure we changed all that much. There was no obvious jump upward in quality, but I think we do pretty good in general and would like to see us continue to hold steady there.
❌ Better guides. We didn’t do all that much with guides. Part of the problem is that it’s a little confusing. For one thing, we have “guides” (e.g. our guide to flexbox) which is obviously useful and doing well. Then there are “Guide Collections” (e.g. our Custom Properties Guide) which are like hand-picked and hand-ordered selections of articles. I’m not entirely sure how useful those hand-curated guides are, especially considering we also have tag pages which are more sortable. The dudes with the biggest are the hand-written articles-on-steroids types, so that’s worth the most investment.
100k on email list. That would be a jump of 35k which is more than we’ve ever done. Ambitious. Part of this is that I’m tempted to try some stuff like paid advertising to grow it, so I can get a taste for that world. Didn’t Twitter have a special card where people could subscribe right from a Tweet? Stuff like that.
Two guides. The blog-post-on-steroids kind. The flexbox one does great for us, traffic-wise, but I also really enjoy this kind of creative output. I’ll be really sad if we can’t at least get two really good ones done this year.
Have an obvious focus on how-to referential technical content. This is related to the last goal, but goes for everyday publishing. I wouldn’t be mad if every darn article we published started with “How To.”
Get on Gutenberg. The new WordPress block editor. This is our most ambitious goal. Or at least I think it is. It’s the most unknown because I literally don’t know what issues we’re going to face when turning it on for more than a decade’s worth of content that’s been authored in the classic editor. I don’t think it’s going to hurt anything. It’s more a matter of making sure:
- authoring posts has all the same functionality and conveniences as we have now,
- editing old posts doesn’t require any manual conversion work, and
- it feels worth doing.
But I haven’t even tried yet, so it’s a don’t-know-what-I-don’t-know situation.
I was thinking about how stage musicians do that thing where they thank their fans almost unfailingly. Across any genre. Even if they say hardly anything into a microphone during the performance, they will at least thank people for coming, if not absolutely gush appreciation at the crowd. It’s cliché, but it’s not disingenuous. I can imagine it’s genuinely touching to look out across a room of people that all choose to spend a slice of their lives listening to you do your thing.
I feel that way here. I can’t see you as easily as looking out over a room, but I feel it in the comments you post, the emails you send, the tweets you tagged us in, and all that. You’re spending some of your life with us and that makes me feel incredibly grateful. Cheers.