iliana

iliana

@iliana

pure of heart and dumb of ass

i work at a computer company and have debilitating adhd. longer intro


see also:


haha.business: 0 current visitors, 646 unique visitors, 775 total page views, 90% bounce rate, 19 second visit duration

As the operator of critical web infrastructure such as haha.business, it is my sworn duty to ensure the website is operating at full potential. I should — no, must — use all the tools at my disposal to consider whether the website is meeting the needs of its many many users.

Here is a picture of the metrics for haha.business over the past month. Metrics are, of course, likely to be limited due to the prevalence of privacy-preserving extensions such as uBlock Origin (which honestly everyone should have installed) and the differences in what filter sets folks use. But our five numbers here paint a useful picture, and tell us:

  1. About 20 unique visitors a day were greeted by the visage of a man in a suit with an awful, noisy tie and an old brick-shaped cellular phone laughing about business.
  2. On average, visitors spent about 19 seconds presumably staring at this man and reading the aliased text surrounding him. It's unclear if this average has outliers removed, so it might be that the average user looks at the page for two seconds, while Business Georg leaves the browser tab open for god knows what reason [author's note: i realize i spent 10 minutes with this open in a background tab while writing this post]
  3. By multiplying two numbers together, we can determine that this website collectively wasted at least 3.4 hours of time on the planet.
  4. Total pageviews is about 20% higher than unique visitors, which probably means something.

What is most interesting to me here is the bounce rate; 90% for a single-page website is a bit puzzling. I would have expected it to be much closer to 100%. Plausible defines the bounce rate as "the percentage of sessions that had one page view only", but the website has one page. This implies that 10% of sessions refreshed the page.

I am choosing to believe that 10% of visitors to haha.business saw a single image and assumed the website was broken. That there was more to see. Perhaps the image changes if you refresh? Perhaps the website will be interactive in some way that didn't load the first time? No, of course not, but I know this having hand-written the HTML.

There is something that has always bothered me about the internet. I have what I think is a decent sense of humor, as do most of the people I surround myself with. And I think it is important to use this sense of humor to good ends; to make the world a little bit weirder. Time and time again, I find people who "don't get" the silly things. It is difficult for me to empathize with them. But having interacted with them over and over again, I'd believe that some percentage of that 10% who refreshed the page looking for something that wasn't there did so because they didn't get it, and wanted to.

I am somewhat curious how haha.business became known at all in some very small parts of internet vernacular. Here's a tip: you can search website URLs on Twitter by just... searching for them. I learned nothing from this search, of course, but it was curious to see the URL end up in a video post on Netlify's Twitter account demonstrating how to do some thing with their service I don't understand. (The website was hosted on Netlify at the time.) More recently, it seems to be used somewhat generically as a way of referring to the meme, often in a "well I sure am doing Business despite being uninterested or unqualified for it" kind of way.

The best part of it all is seeing people wondering who owns it; a handful of tweets ask a question I dare not answer several months later. I am perhaps outing myself as the sole maintainer of a mysterious corner of the internet, but so far at least several dozen people know it's me, and it hasn't gone much farther past that. I think that's interesting.

I leave you with one more metric: Multiplying the total size of the page — 146.6 KiB, according to PageSpeed Insights — by the pageviews nets me about 111 MiB of data transfer. That's only slightly more than a 100 MB Zip disk from 1994!


in reply to @iliana's post:

yeah Plausible's "bounce rate" statistic is. one of those things where if i were Doing Business i would probably care a lot about it but since i am Yelling With A Bullhorn At A Street Corner i don't really give a shit

Professional web analyst here. Web data is fun because it's so janky! Many things that seem odd start to make sense when you consider the myriad ways that the data can be broken. Two ideas spring to mind for the bounce rate.

  1. Pre-loads/pre-fetches. Some web clients (and websites) will begin loading a linked page before a user clicks a link. Sometimes this is for responsiveness, in a handful of cases it's so the user can preview the website before they click. Some of these previews are counted as pageviews. I remember metrics from Safari turned screwball a few years back because of link previews.
  2. "What is a session, really?" Almost everything interesting in web data is defined in terms of sessions, but sessions are defined in terms of vague hand-waving heuristic bullshit. It's the least-well-founded of the foundational measurements. Say a person views the page and leaves. Then twenty minutes later they decide to share it with a friend, so they re-open it and copy the URL. Logically that's two bounced sessions. But most tools will report that as a single session that's twenty minutes long! And for Google Analytics, it could be either depending on whether the second visit is via Google search or re-opening from history!

Plausible Analytics in particular also uses an imprecise method for measuring Unique Visitors (it's more privacy-friendly, although I strongly doubt it avoids any obligations under GDPR/PECR). If two users happen to share an IP address and user-agent, then Plausible will count them as a single user. This is uncommon in personal settings, but people Doing Business are likely to have the same laptop config and using either the same office or VPN. So if people share this around an office, that will inflate your pageviews-per-visitor.

wow! thank you for this deep dive into your very official business webbed site metrics. i have learned a lot about how to interpret these numbers and what to watch out for – bounce rate was quite surprising, and the "collective time wasted" metric was very enlightening! great learnings all around 🙏