How do Google and Bing compile search results?

You might not be familiar with the concept of proprietary eponyms, but the chances are you use them every day.

If you do the hoovering instead of the vacuuming, or ask people for a Kleenex instead of a tissue, you’re using a brand name to stand for a generic action or item.

Sometimes, there are obvious alternatives to eponyms. It’s not just Velcro who make sticky fastener pads, and you can refer to non-Jeep-branded vehicles as 4x4s, SUVs or off-roaders.

However, it’s harder to think of an alternative to Google – the ultimate proprietary eponym.

Nobody talks about Binging something, even though Microsoft’s Google rival currently holds almost five per cent of the UK search engine market.

We rely on search engines constantly, particularly when we have more free time. And looking back at our search histories can be amusing, embarrassing, nostalgic and depressing.

Yet few people appreciate the complexity of compiling Google and Bing results, or the ways in which these search engines define one webpage as being more relevant than the next.

Search and ye shall find

In the internet’s formative years, search engines were effectively advertising directories. Companies had to pay to appear in Yahoo’s listings as recently as 2002.

However, rival search engines had already recognised the merit of providing listings based on relevance rather than revenue.

Consequently, companies like Google and Microsoft developed complex mathematical algorithms, which attempted to rate different sites in order of significance.

Early Google and Bing results relied on factors like the number of times a particular search term appeared on the page, which quickly inspired a concept known as keyword stuffing.

Although it was hard for people to read the resulting content, search engines prioritised it because they felt it was highly relevant.

As users began complaining about low-quality content achieving high ranking scores, the search engines refined their algorithms and keyword stuffing fell from grace.

Today, such attempts at gaming the system are collectively known as black hat marketing. Websites adopting these unsavoury tactics are downgraded in rankings, or blacklisted.

Key factors in website ranking

The exact composition of search engine algorithms remains a secret, to prevent unscrupulous black hat marketing agencies trying to cheat their way onto page one.

(Since the vast majority of users don’t click onto the second page of results, a page one position is critically important.)

However, we do know a great deal about the general principles underpinning search engines. These attributes help to determine where a website will perform in Google and Bing results:

• How long the site has been live
• How much traffic it receives, and the average time visitors spend on each page
• How quickly it loads, which affects the user experience on slower connections
• The number of links to that page or site from third-party websites
• Whether on-page content is obscured by floating text boxes or pop-up adverts
• The use of long tails – three or four-word search strings like ‘broadband deals UK’
• Whether the page contains headings, captions and a brief description of its content
• The top-level domain – a co.uk TLD will perform better in Britain, but worse abroad
• The prevalence of internal and external links to other webpages.

It’s also worth noting there are differences in how Google and Bing results are compiled.

Bing’s search engine analysis focuses on the first 100 kilobytes of each webpage, making headlines and opening paragraphs more important.

Bing also prioritises large numbers of social media links to a website, and the presence of multimedia content.

This is the search engine powering Amazon’s Alexa smart speaker system, so websites written in conversational English are fairly likely to rank highly in Bing results.

Conversely, Google is more interested in how when a site was last modified, as well as the volume and quality of inbound and outbound site links.

HTTPS websites perform far better in Google results than those without secure connections, and it’s hard to overstate the SEO benefits of regularly uploading original content.