Or rather a change of pace, a change of content and a change of design.
So far, I’ve mainly written articles with the primary intention of presenting “the facts”, even if you do have to distil them from the murky casserole of terrible jokes, swearing and catpix. These kinds of posts might be about a pharmacological process, a recipe, tek, or perhaps an update on a news story or developing situation, and I’ve tried to write something about once every week or two. The problem is, these kind of posts usually require all sorts of crazy shit like “forethought” and “reading stuff”. It’s not that I’m incapable of such bizarre activities, it’s just that I’ve often got more to say than the time it would take to say it properly. I’ve always felt it necessary to try and cover as much of a topic as I can, while trying to present it nicely as possible, but that level of effort means I can only be arsed only have the time to write articles like that sporadically. As a result, I’ve ended up not bothering to write about a number of interesting topics that perhaps weren’t that big of a deal because there’s not enough time in the day to cover them all “properly”. So, I quit.
Hah. Only joking.
I’m going to try blogging a bit more casually instead. The plan is to write shorter posts, perhaps not going into all the details, or considering all eventualities, or whatever. I might be more likely to just quote the odd article, for instance, or maybe just write down my opinions on something. The upshot is I’m hoping to post more frequently about a broader range of subjects, so everyone’s a winner! That is, unless you hate what I have to say, in which case, why haven’t you fucked off yet?
I’m not giving up writing longer posts altogether though. Far from it! In fact, writing more posts in general will probably encourage me, and I’ve had a few topics in mind for ages.
The next thing I’m changing is the blog’s design.
No, nothing that amazing, don’t panic, just a few tweaks here ‘n’ there to find what works best for the largest amount of people. And I’m going to do it with science!
For a while now, I’ve been thinking how long the blog takes to load, especially the front page, and especially for new users with an empty cache. Shaving off a second or two doesn’t sound like a big deal, but you’d be surprised on how much of an impact the speed of a website can have, but it affects almost everything, such as how long people stay on the site, how many pages they view in that visit, how likely it is they’ll pass your link around, how likely it is they’ll buy something and how much they’ll spend (for ecommerce sites) and potentially where the site will rank in Google.
Have a look at some more graphs ‘n’ charts to see just how much those few seconds matter.
I’ve already gone through the entire blog and compressed the fuck out of every single image, shaving off an entire megabyte in total, but only a fraction of those images are on the front page. Luckily, there’s a few things I can do that will specifically affect the front page:
- Display less posts
- Display only the first bit of posts (with a “Read More” link to the rest)
- Don’t display as many images
- Write shorter posts
That last one I’ve already covered, so what about the first three design-related options? Obviously, they’d all reduce load time, and, if I went as far as displaying only the one post’s headline and first three letters with zero images, load time would be minimal, but then no one would stick around anyway. A site needs to be fast, but it still needs some content to keep people engaged.
To find the most effective trade-off, I’ve set up a little experiment… In the marketing biz, it’s called a split test, and it works something like this:
You create several different versions of something — it could be an entirely different page layout or design, or it could be as simple as placing a little button in different places on the same page — and then you randomly assign each visitor to one particular variation (making sure they always see the same variation on subsequent visits). Then you just sit back and watch the data pour in.
After a while, you can then start to compare how each group of visitors behave. Version #3 might keep people on the site for twice as long, but version #2 might see people return thrice as often. When enough data has been collected, hopefully a victor will emerge.
Yesterday, I created six versions of the front page and got the experiment under way. Here are the details:
|V1||10 full posts are displayed — this is how it’s been for everyone up to now, so this is the control|
|V2||8 full posts|
|V3||6 full posts|
|V4||Latest two posts displayed in full, next two are max 1300 characters, next two are max 800, rest are 500 max — 15 posts in total|
|V5||Same as above, but 20 posts (so more of the 500 character excerpts at the bottom)|
|V6||20 posts again, but only first post in full, next two 1300 characters max, next five 800 max, rest 500 max|
Anything that’s not a full post won’t have any images in it either, and older pages (there’s a link to older posts at the bottom of the front page) will display the same as the last post on the front page, so older pages of V1-3 will be exactly the same as the front page, while older pages of V4-V6 will have 15 or 20 posts of 500 character excerpts with no images. Also, the excerpt length is approx, and will (hopefully) round off to the nearest whole word.
When it’s clear that one has performed better than the other, I’ll stop the experiment and implement the winning version for all users. Then I’ll probably test something else. Of course, there’s no guarantee one of them will perform any better, in which case, I’ll pick the one I like best. Whatever happens, I’ll let you know, probably via some off-the-cuff scribble of a post.
… Which is what I was hoping this post would be, but it looks like I’ve ended up writing loads again.