Bloat-to-content ratio is killing the web - here's how to fix it

The publishing model is broken by horrible practices but there is a way out

An amazing recent blog post from Mozilla developer Les Orchard shines a bright light on one of the Web’s darker habits.  Despite its title, it’s not a particular pop at The Verge, which the author takes pains to say he likes. It’s a look at the way revenue models for web content sites in general have utterly poisoned the very thing they depend on.

In the post, Orchard looks at a one-page Web article. This has around 75 kilobytes devoted to actual content - words someone has written and the HTML wrapping they come in. But to read them, he had to download in the region of 10 megabytes - more than a hundred times as much - of trackers, adverts, Javascript and the like. As Morchard points out, that’s a hell of a chunk of change out of his data plan, not to mention a massive overhead on loading time, browser responsiveness, power sucked out of batteries and other things entirely inimical to the mobile experience everyone swears they want to make better.

As the raw text itself stripped of its formatting is 1,500 words taking around 10 kilobytes, there’s actually a 1000:1 ratio of bloat to content. Something like 40 companies have hitched a ride on that one page through embedded trackers, adverts and the like, and they all want a bit of you.

Put it this way - those modest 1,500 words have to support a thousand times their own weight, and any revenue generated from the reader - you - gets split up to support the entire writhing mass.  No wonder the predominant model of making money from internet publishing is to lose it.

Morchard goes on to talk about various models that might make things better, such as moving advertising to the browser for insertion, and to be sure there are engineering solutions that could cut out a lot of the heavy lifting. But that’s no real fix. The answer isn’t to have better ways to deliver more advertising. It’s about how to deliver less.

Because that’s what we all - as human beings, rather than ad sales robots - want. Think back to the golden age of newspapers: a solid page of text with tens of actual stories and pictures, with perhaps three or four discreet adverts tucked around the edges. And those adverts cost the advertisers serious money because they were getting a unique chance to present their case to an intent readership. The readers got their content hit, the advertisers got results, the publishers made money. Everything that’s now been broken by treating content as homeopathy, molecules of goodness lost in litres of sugar water.

There is a way to fix this and get back to that working model. It’s simple, and it can change everything. It’s called search.

Search engines - OK, the one search engine anyone uses - ostensibly work by sorting the billions of Web pages for relevance to the reader. Google’s mighty algorithm grew by cleverly assessing actual content through the number and quality of its links to other quality content. As a result, the vast majority of web traffic is now driven through Google. But what Google cannot do is downgrade pages because of the quantity of legitimate advertising and marketing dross around that content. Google makes its money from delivering advertising: if it targeted its competitors directly, it would be in court for the rest of its (surprisingly short) life.

Excellent. That leaves the market open for a new search engine, one that follows the same basic rules for content quality but also aggressively and unstintingly down-ranks pages on the bloat-to-content ratio. Load up your page with a thousand bytes of marketing to one byte of content? Nobody will ever see it. Have a modest page with one advert for one piece of work? You’ll get seen. What to call a search engine that delivers decent core content with minimal wrappers? Quality Street seems suitably festive.

Not only would Quality Street deliver stuff in a format that readers (and writers) desperately want and need, it would be immune to any Google countermeasures. What little advertising it would tolerate would be delivered to readers who actually cared about what they were reading, which is catnip to advertisers. That means actual money, which can be used - shockingly - to build even better content, rather than ever more baroque ways to prop up a dying idea.

At a stroke, it could create a situation where any content site that actually put its readers first would have an automatic advantage, and restore the idea that quality matters. It would let readers - you - take back the Web on your own terms, without asking permission from anyone.

Which is the idea that gave birth to the Web in the first place. This time, it might even work.