You tap a link. For a moment the screen is white, then the headline lands, then a paragraph or two. So far, so good. And then the page begins to settle — a strip appears above the text and pushes it down, a box opens on the right, something starts playing quietly, and the sentence you were halfway through is now somewhere near the bottom of the screen.
Nothing has gone wrong. That is simply what loading a modern page looks like from the outside. Underneath, the page you asked for was only the first of a long series of requests, and most of the rest were never really addressed to you.
The page you asked for is only the first request
When you open a web address, your browser fetches one document: the HTML. That file is short. It is essentially a list of instructions — here is the text, here is where the pictures live, here are the styles, and here are the other addresses you should go and collect before this looks like a finished page.
Your browser reads that list and starts fetching. Images, fonts, stylesheets, scripts. Some of those live on the same site you visited. Many do not. A single article page routinely pulls resources from a dozen or more separate hosts, each one requiring its own connection, its own handshake, and its own moment of waiting.
The important detail is that these requests are not neatly queued behind one another. They cascade. One script arrives and asks for two more. Those two arrive and ask for a measurement, an identifier, a creative file, a fallback creative file. The browser obliges, because obliging is its job.
This is why a page can be readable in half a second and still feel unfinished ten seconds later. The reading part finished early. The fetching part did not.
The article was ready almost immediately. It just wasn't allowed to sit still.
A short tour of what else is on the guest list
If you have ever opened a browser's developer tools and watched the network panel fill up, the shape of the problem becomes obvious very quickly. Beyond the article itself, a typical commercial page tends to invite a fairly predictable cast.
Display advertising. Not one file, but a chain: a slot definition, an auction, a winning creative, and the assets that creative needs. Each link in the chain waits for the one before it.
Measurement and analytics. Scripts that record which sections you saw, how far you scrolled, where your cursor paused, and how long you stayed. Useful to the site. Invisible to you.
Tag managers. A single script whose entire purpose is to load other scripts, decided at the moment you arrive rather than when the page was built.
Social and embed widgets. Share buttons, comment systems, embedded posts and players — often carrying their own frameworks along with them.
Consent and preference tools. The banner itself, plus the logic that decides which of the above are allowed to run and in what order.
None of these is sinister on its own. Collectively, though, they are the difference between a page that arrives and a page that keeps arriving.
Why the page keeps moving under your thumb
The most irritating symptom is not slowness. It is movement. You start reading, and the paragraph slides away from you.
This happens because many of those late arrivals need space, and the page does not always know how much space in advance. A slot is declared, the auction runs, and only then does the browser learn that it must reserve a rectangle of a particular height. Everything below that rectangle moves down to make room.
On a desktop screen with a mouse, this is annoying. On a phone, where your thumb is already hovering over a link, it is worse: the thing you meant to tap moves, and something else takes its place a fraction of a second before your finger lands. Browsers and standards bodies have spent years trying to discourage this pattern, and it is measurably better than it was — but on pages with a lot of late-loading content it has not gone away.
Autoplay, sound, and the prompt you did not ask for
Two arrivals deserve their own mention because they interrupt rather than merely delay.
The first is video that begins on its own. Even muted, an autoplaying player is doing real work: fetching a stream in segments, decoding frames, and drawing them continuously for as long as it is on screen. It is usually the single most expensive thing on a page you never chose to watch.
The second is the notification permission prompt — the small panel near the address bar asking whether a site may send you messages later. Accepting one is easy, and the consequence arrives days afterwards, from a site you no longer remember visiting, in the corner of a screen you were using for something else entirely.
Both are reversible. Browsers let you set notifications to ask-quietly or block by default, and both major desktop browsers now restrict autoplay with sound. The prompts themselves, though, keep coming — because asking costs a site nothing.
The part your device quietly pays for
Every additional request has a cost that does not appear on the page. Each connection consumes a little data. Each script must be parsed and executed, which means processor time. Each animated or video element means the screen is redrawn far more often than reading text requires.
On a fast connection and a recent laptop, most of this disappears into the margins. On an older phone, on a train, on a metered plan, it does not. The page takes longer, the device runs warmer, and the battery figure drops faster than the time you spent reading would suggest.
It is worth saying plainly: this is not a fault in your device, and it is rarely a fault in the site's own writing. It is the accumulated weight of everything the page was asked to bring along.
Which is also why the fix is rarely a faster device. The load has to get lighter.
A few habits that lighten the load
Set notifications to blocked by default. In Chrome this lives under Settings, Privacy and security, Site settings, Notifications. You can still allow individual sites you genuinely want to hear from.
Close the tabs you finished with. A background tab with a live player or a polling script is not idle. It is still working.
Prune your extensions occasionally. Every extension runs on the pages it has permission to see. Two you use are lighter than nine you have forgotten about.
Clear the cache when a single site misbehaves. Not as a routine ritual — a warm cache is what makes repeat visits fast — but as a targeted fix when one page renders oddly.
Use reader mode for long articles. It strips the page back to text and images, which is often exactly what you wanted in the first place.
Consider a content blocker. This is the one that changes the shape of the problem rather than working around it, because it stops the requests instead of waiting for them.
Where the Total Adblock browser extension fits
A content blocker sits between the instruction list and the fetching. When the page says "now go and collect these forty things," the extension checks each address against a set of filter lists and declines the ones that match known advertising and tracking endpoints.
The practical effect follows from that. Requests that are never made cannot be slow. Rectangles that are never reserved cannot push your paragraph down the screen. A player that is never initialised does not decode video in a tab you stopped looking at.
Total Adblock is a browser extension for Chrome that works this way. It blocks advertising and many common trackers, suppresses unwanted pop-ups and notification prompts, and includes video-ad blocking on major sites in its Premium tier. It installs from the Chrome Web Store in a couple of clicks and runs quietly from the toolbar, with a per-site toggle for the pages where you would rather leave everything on.
It is worth being clear about scope. An ad blocker is not an antivirus, not a VPN, and not a general performance tune-up for your computer. It does one narrow job — deciding which requests a page is allowed to make — and that narrowness is the point.
