Despite my internet connection being literally thousands of times faster than my old dial-up connection was decades ago, it often feels like websites still take just as long to load.
The obvious reason for this is that as the size of the old internet pipe has gone up, so has the volume of data that makes up a typical website. Even a basic browser website like Wikipedia takes up a decent chunk of memory, for no obvious reason on the surface. So what’s going on?
From Lightweight to Bloated
When all you had was (in a perfect world) 56Kbps of bandwidth, your website has to be lean and mean. Early sites were kilobytes in size, and were mostly text. When images were there, they were low-res JPEGs with aggressive compression. I once downloaded a 64MB music video on my dialup modem, and it took an entire weekend to finish. So video embedded on a site? Forget about it.
It’s that age-old paradox that also seems to hit real-world highways. No matter how many roads, bypasses, and flyovers you build and how many lanes you add, you’re always getting traffic jams. It’s probably the same psychology and economics that drive the weird phenomenon of energy-saving lights. Instead of reducing our power bills, we just add more light for the same energy cost!
Culprit number one is simply richer media. We have high-resolution screens on all our devices, and so we need high-resolution images if we want a site to look good. Modern image formats like WEBP have gone some way to reducing file sizes while keeping the quality high, but the overall trend is more images with more pixels, and that bloats things.
If it was just still images, that would be one thing, but a typical website has auto-playing videos, audio, animated GIFs, and plenty of rich moving media built on the technology of the website itself, such as HTML5.
Frameworks, Libraries, and Code Bloat
A modern website isn’t something that a person sits down and codes out in raw HTML like the old days. Just like coding an app, web developers rely on vast libraries and sophisticated tool to help them create these sites quickly and efficiently. However, this does lead to a situation where sites are bloated by reliance on these libraries and frameworks, since the site developer calls on them even for relatively simple things.
Over time, sites can just accumulate code as it changes and updates. Old code, unused CSS, outdated plugins, and debugging scripts are left behind. Which doesn’t affect the site’s functions, but does eat your bandwidth.
Advertising, Tracking, and Analytics
When it comes to making sites richer experiences, for users, then there’s some justification for the bloat, but what about when it’s just to make someone richer? Well, the truth is, of course, that websites cost money to create and run, and they need to make money. Unless users are paying for those services directly, the primary way to sustain a website is through advertising.
In the early days of the web, a site might have had a single banner advert at the top and then again at the bottom of the page. Endless scrolling didn’t exist yet, which is another reason for bloat by the way!
Now there are entire advertising systems built into websites which, again, are only there because, in general, people are seemingly unwilling to pay for content even if they find it valuable. However, there’s no denying that ads, trackers, and analytics weigh websites down. Necessary evil or not.
The Feature Creep Problem
The last big reason for all this bloat, at least in my opinion, is the extreme feature creep that’s taken hold of websites. In the old days, sites used to be a static page of information, but now it’s like sites try to be every type of web app for every user.
Chat widgets pop up when you don’t want them, there’s software running in the back watching everything you do, and notifications and more notifications. When every website tries to build a TikTok or YouTube-like elements into it, it doesn’t work, and then it’s not cleaned up properly later, you end up with a graveyard of failed “improvements.”
Why It Matters
We like to complain when our software isn’t optimized, because it means spending money on faster processors and more memory just to stay in one place. So why not level the same complaint against websites, which are effectively now software applications delivered live via the web? This isn’t just an issue about confusing, ugly websites or long load times either.
Bloated websites eat data on capped connections, congest the internet for everyone, burn more energy, and increase hardware costs. The web keeps getting heavier not because it has to, but because it can. Faster pipes have made developers complacent, letting complexity sprawl unchecked. It will take a reset to performance as a core design value to make things better, but would you accept a more stripped-down web?