What is a ranking factor?

I decided to write this for a couple of reasons. One is that I’ve seen a lot of potentially misleading Tweets on the subject recently (naming no names!), and the other is that it’s related to another pet peeve of mine, about ranking factor studies.

What is a ranking factor?

A ranking factor is a variable that a search engine uses to decide the best ordering of relevant, indexed results returned for a search query.

Note that I’ve said the decision is between relevant, indexed pages – a good illustration of this distinction is the often absurdly high number shown beneath your query when you perform a Google search, such as the 643 million “potatoes”-related pages shown here:


Most of these pages are not particularly relevant, but this is the set that ranking factors are seeking to order. Some of the factors used to establish what is relevant to include in this list are also ranking factors (for example, having links with the anchor text “potatoes”), but they are not the same thing.

Perhaps the most famous ranking factor is Google’s PageRank – invented at a time when, proportionately, a great deal more web browsing was done by clicking links from popular pages, the role of PageRank was to approximate the popularity, and therefore, by extension, authority, of a page on the internet.

How are ranking factors combined?

A search algorithm might take ranking factors like PageRank and weight, sum, or multiply them in any way seen fit. The objective is to combine them in a way that achieves the “best” results – for example, presenting the results that users are most likely to click on at the top. According to this CNBC interview from late 2017, metrics Google might optimise for include time to SERP interaction and rate of bouncing back to search results.

Amusingly, probably the best insight we’ve had into how Google combines ranking factors came from a question about featured snippets, asked to Google’s Gary Illyes by Jason Barnard earlier this year. I say amusingly because Gary seemingly had no need to give such an in depth answer to this question, but the model he describes is fascinating to explore and extend upon – you can read more about it in Jason’s article here.

What ranking factors are there?

We don’t know! In fact, we don’t even know how many there are. Probably a great many. (Although, not everything that influences rankings is a ranking factor – more on that below!)

Google occasionally explicitly confirms a ranking factor (like HTTPS, or page loading times), but often this is as much as anything to push the SEO industry to change the internet in a direction they’d like.

It’s in their interest to keep actual ranking factors and their relative importance very close to their chest, as their algorithm is part of their advantage over their competitors.

Is user experience a ranking factor?

Not exactly, no – “user experience” is not a metric. However, we do know (for example, from that CNBC article referenced above) that many of the things Google is looking for correlate with a good user experience. We also know that Google highlights things like slow loading times or excessively small font on mobile devices as SEO issues, which suggests an interest in this kind of factor.

There’s an ongoing debate about whether Google tries to measure this directly (for example, by looking at the click through rate of specific search results, compared to what might be expected, and adjusting their ranking on the fly), or whether they merely adjust their algorithm to look for things that correlate with user experience improvements. Many real world experiments suggest the former, but Google’s official line is the latter.

It’s in their interest to keep actual ranking factors and their relative importance very close to their chest, as their algorithm is part of their advantage over their competitors.


Misconception 1: All metrics that correlate with rankings are ranking factors

This misconception is probably in some small part my fault – and also the fault of others who, like me, have published so called ranking factor studies. These are actually correlation studies – we can look at what qualities are typically held by well-ranking pages, but this does not mean that those qualities are necessarily ranking factors. Facebook Likes, for example, correlate well with Google rankings. There are a few possible explanations for this:

  • Sheer coincidence (unlikely – that’s the point of statistical significance thresholds – but possible)
  • Pages that rank well tend to be seen a lot, therefore end up getting a lot of Facebook Likes (i.e. the causation is in the opposite direction to Facebook Likes being a ranking factor)
  • Pages on popular websites receive both many Facebook Likes and strong Google rankings (i.e. there is something else that causes both, rather than one influencing the other)
  • Google assesses the number of Facebook Likes that results have, and takes this into account when ordering them in search results (i.e. Facebook Likes are a ranking factor).

Misconception 2: A ranking factor is anything that causally affects rankings

If you engage in link building, you are engaging in an activity that is designed to influence a ranking factor – something you believe Google will consider directly in their algorithm. However, there are lots of things you can do to make your site rank better, but which are not in themselves designed to influence ranking factors.

Perhaps my ultimate claim to fame is that I used to work at the UK’s busiest Little Chef (a now-defunct chain of roadside grills), where I was a cook. If my boss had sent me on a training course, this may have resulted in a few things that could go on to improve the business’s rankings in Google Search, such as:

  • Glowing coverage in local press, due to the restaurant becoming known locally for its excellent food
  • Bloggers mentioning and linking to the Little Chef website, having been pleasantly surprised at their experience
  • People clicking on the site even when it’s position 8 in the search results, because they know and love the brand, having had so many wonderful meals there

Sadly, my boss did not send me on a training course, and Little Chef eventually died, their premises being ignominiously absorbed into the empire of Starbucks. However, that does not mean that “kitchen staff competence level” is a ranking factor. It is not something that Google is attempting to directly measure and include as a variable in its algorithm, therefore it is not a ranking factor.

However, if you are running an ailing roadside grill, and you wish to improve your rankings, you could try having competent kitchen staff. The causal link is there, even if the ranking factors are involved only indirectly.

Misconception 3: Ranking factors are dead / don’t exist

I’d be the first to say that ranking factors may not be a helpful for SEOs anymore. In fact, I’ve written about it on this very blog.

Because ranking factors are so many and so unknowable, it’s often better to aim for what Google is optimising for, therefore avoiding the need for any Kremlinology.

However, that does not mean that ranking factors are not a thing – they are still a crucial part of how search engines work at a very basic level and understanding the theory gives you a solid foundation to your knowledge.

Misconception 4: Bounce rate, time on page, and/or conversion rate are ranking factors

These are metrics in analytics, which means Google does not have access to them – even if you use Google Analytics. Furthermore, they’re easily manipulated, and often don’t obviously correlate with good outcomes.

For example, Google might be interested in the rate at which I return to search results after clicking on your site – this would indicate I was unhappy with that result. However, they can get this information directly from their own analytics on search results, and the bounce rate in your analytics could be misleading – if I read your page, get the answer I want, and move on in my life, that’s a good search result from Google’s perspective, but probably a bounce in your analytics.