In December 2014 I decided to post every day for a month. I had two goals: get back into the habit of daily writing and establish a baseline “active month” for Nuketown in Google Analytics. My secondary goal was to increase traffic to the website (both year over year and month over month) but I wasn’t sure what percentage was reasonable, but I wanted to see increases in unique sessions, unique users, and unique page views.
I succeeded on all counts. Content-wise, I added a new post to the site every day in December. Four of these posts were re-prints from my days writing for the now-defunct GameCryer.com, but everything else was new for December. I also made a handful of microcontent updates (the pink-colored blocks on the homepage that also get posted to Twitter).
It wasn’t easy — there were a few nights where I didn’t start writing until 10 p.m., and snuck in the update just before midnight, but I did update the site every day.
Traffic to the website increased nicely. Compared to the same time last year, sessions were up 30%, users were up 22%, and page views were up 43%. Adsense revenue, as sad as it is, also saw a 30% bump. Here’s the full chart:
Year over Year (2014 vs 2013)
|Pages / Session||9.66%||1.64||1.49|
|Avg. Session Duration||-6.07%||00:01:16||00:01:21|
|% New Sessions||-6.19%||81.54%||86.92%|
Traffic was also up on a month over month basis, with page views in particular jumping 60% versus November, and session duration jumping 70%.
Month over Month (December vs. November)
|Pages / Session||16.72%||1.64||1.40|
|Avg. Session Duration||70.96%||00:01:16||00:00:44|
Note these numbers don’t include my own authenticated traffic to the site, because Google Analytics isn’t tracking Nuketown’s logged-in users. It does include my own spot checks of the unauthenticated views.
This wasn’t the first time I’ve tried a post-a-day project. Back in April 2014 I did my 30 for 30 project, in which I was shooting for simply updating the site (posts or microcontent) every day for 30 days. It didn’t have nearly the same impact on site traffic; during that effort page views only went up about 3%.
The Top 25
The 25 top stories for the month are interesting. The Top 5 most visited pages are dominated by the homepage and several evergreen pages — pages that consistently get a lot of traffic regardless of the month.
The 10 evergreen posts in the top 25 accounted for 1,806 page views (32% of the total for the month); the 15 new posts accounted for 1,097 page views (18% of the total for the month. The most popular of the new posts (anything north of 100 page views) dealt with Dungeons & Dragons. The list is biased towards RPGs in general simply because that’s what I’ve been writing about the most, but I was glad to see my “Battle of the Five Armies” review, my “Winter Reading List”, and the “Top of the Pile” column made the list (even if their traffic was, well, anemic).
In the Winter of 2009…
All of this is a far cry from where Nuketown’s traffic used to be 5 years ago. Sessions are down 62% (3,413 vs 9,186), users are down 64.12% (2,871 vs 8,002), and page views are down 60.38% (5,592 vs 14,115).
Why? Looking back at the top 25 pages in December 2009, it’s pretty clear: some of Nuketown’s old evergreen posts … wilted. Pages like “Installing Ubuntu Linux on a G3 iMac”, “Linksys, Open NAT, and Halo 3”, and “Installing HP 1020 Drivers on Mac OS X” used to be terribly popular. Those three pages alone generated 2079 visits, but the problems they solved probably aren’t issues for people in 2015.
The problem was that these pages had high bounce rates (generally around 90%) — people visited, then they left, without ever exploring the rest of the site (hardly surprising given that the rest of the site, while very geeky, is not dedicated to debugging drivers and network issue). That said, at least 10% of those people did venture deeper into the site, and losing these high-traffic pages has led to an decline in overall traffic to the site.
This is all part of the natural ebb and flow of the website — things get popular with Google, they generate traffic for a few months or years, and then they fade away. My hypothesis is that my general lack of blogging on the site has prevented new Google-friendly niche pages from being born. I’ve solved just as many weird technical problems between 2009 and 2014 as I did between 2004 and 2009 … I was just far less likely to write about them.
So where does this all leave me? Nuketown is a hobby. I do it because I enjoy writing, and it gives me an outlet. It’s also a perpetual experiment where I can try out new web designs and approaches to content strategy (which is helpful for my day job as the director of web development at a small college). I don’t rely on Nuketown to pay the bills, though the occasional Adsense check does let me buy a new video game, RPG, or novel from time to time.
I’ve contemplating giving up Nuketown, but never seriously. Yes, it’s a sizable time investment that doesn’t yield buckets of cash or world-wide acclaim … but it helps keeps me sane.
Going forward I’d like to play with the update formula to see if I can find the sweet spot between the fast-and-furious post-a-day pace and the anemic “wait, I last posted three months ago?!” pace from earlier in the fall. Right now I’m thinking that updating three times a week is reasonable — it keeps up a steady stream of content, but doesn’t keep me up until midnight writing. This also helps me up the quality of my writing by giving me at least a day to proofread what I wrote (always a challenge for me, all the more so now that I have much less free time at 43 than I did when I was 23 … or 33).
I have no interest in chasing page views so I’m going to keep writing about the things I’m interested in. With a little luck, some of those will lead to a nice bump in traffic, and a few more folks will join the ol’thermonuclear burg’s community. And if not? Well, that’s ok too. Nuketown is meant to be fun; as long as it hits that goal, it doesn’t really matter how many visits it gets.