There’s long been a debate about how much the quality of a Web page’s HTML affects its performance in search engines.

The high search engine results gained by many poorly coded Web pages are often cited as evidence that search engines don’t really care about the quality of code. On the other hand, developers of standards-compliant websites tend to believe that properly coded pages perform better in search engines.

It’s true that many Web pages with bloated, slapdash HTML do perform well in search engines – sometimes getting top results. However, this only proves two things. First, that the quality of the code is only one factor in a page’s search engine success and that second, in order to deliver a relevant set of results, search engines have to cope with Web pages that are coded to – ahem – varying standards.

Interestingly, Google has recently made clear that it now takes into account a page’s load time as part of its overall quality assessment of a page. That information is now used by Google to create a quality score – which directly affects how much a company pays for its AdWords clicks.

This is probably the first time that we’ve seen a direct financial incentive for making Web pages as lean as possible. The system used is – typically for Google – quite sophisticated.

It tries, for example, to ensure that loading issues caused by geography (or, more accurately, caused by the different speeds of each country’s networks) isn’t a factor in giving a quality score. So, if your site is in the UK, it is measured against other sites in the UK, and so on. You can still host a site wherever you want, without that being the cause of a search engine penalty.

What the system boils down to is that Google believes that a “high-quality landing page” should have a “fast load time as well as feature unique, relevant content”.

We know that Google is already pretty good about assessing how unique a Web page’s content is – Web pages with copied or syndicated content perform significantly worse than those with unique content. (If this distinction wasn’t made, then everyone’s search results would be cluttered with hundreds of different pages with essentially the same content.)

Now it seems that Google is assessing the quality of a page too – including its load time. At the moment, Google is only looking at the speed of loading the HTML – but is clear that, in the future, the load time of other elements, such as Flash movies, video, JavaScript and other such technologies, will also be considered as part of the quality score.

Also, Google is (at least publicly, since it publishes little detail on its algorithms) only using this to grade the quality of pages within its AdWords system. Speaking diplomatically, Google doesn’t say that poor-quality pages are penalised – just that high-quality pages are given a higher rating, which leads to lower cost minimum clicks.

It doesn’t take the brightest button in the drawer to see where this is leading, though – that Google will include the use of a quality rating within its free search results as part of main search algorithms. Perhaps it is already doing so, to some extent – we’ve seen on its accessible search that it is able to deliver higher results for better-coded sites.

This is going to be seriously bad news for some websites – and for some leading content management systems. Slow pages annoy visitors – and now they are going to be found wanting by Google too.
But for the Web as a whole, especially in the longer term, this is going to be a good thing – providing a real financial incentive (though possibly more stick than carrot) to deploy Web pages that are well-coded, lean and standards-compliant.