Gone are the days of keyword stuffing, writing 300-600-word blogs. Thank god.

Gone are the (far more recent) days of ‘exact phrase’ matching, and, although it has been spoken about for many years now, we really are in a place where content must truly be created for the user. Actually, let me re-phrase that – not just created for the user, but created using the guidance of the user. New algorithms are geared towards prioritising content that meets the search intent of the user.

In the current Google climate, by NOT taking this approach, you will see your ranking affected in Google’s search results – if they haven’t already been impacted by the change.

In its most simple essence, Google exists to provide its users with the most relevant, informative and valuable answer to their search query. One way it can do this is by crawling the content, looking at the language and tonality of the content, its structure and formatting, and the quality of the content.

At this point maybe I should add a caveat and say this is one of the APPARENTLY INFINITE ways Google ranks your website and content.

Another ranking factor Google adopts is user behaviour. This makes sense, right? How better to decide if this content or website is of use or value than by monitoring the way users interact with it?

So, what has Google said?

For a long time, not a lot. There seemed to be inconsistency from Google HQ, but for several years now there has been a few nods towards the fact that user behaviour does hold weight.

Which user behaviour metrics are ranking factors?

Click-through rate and immediate post click behaviour (including dwell time)

In this section we look at four different events that can happen after a user finds themselves presented with the results to their search query.

Of course, click through rate is obvious – a user is presented with your result and they click through. Click-through rates are essentially a ‘user vote’; the user believes that this result is going to answer their search query. The data collection doesn’t stop there, the amount of time they subsequently spend on the page (dwell time) is also monitored and analysed.

In fact, Google has previously spoken out about how they use click through data, stating:

‘The ranking itself is affected by the click data. If we discover that for a particular query, hypothetically, 80% of people click on result no.2 and only 10% click on result no.1, after a while we figure probably result 2 is the one people want, and we’ll switch it.’

So, lets look at four example of user behaviour and what they signal to the search engine.

1. A user clicks on your listing in the SERPs, but quickly hits ‘back’, clicking on the next link. This sends the message that they don’t feel they need to amend their query, but instead, your result didn’t provide them with the information they were looking for. This will send a ‘negative relevancy’ signal to Google suggesting that when it comes to this query, you aren’t delivering value.
2. The second scenario is that the user clicks on your listing and stays on the page for some time, before returning to the SERPs and clicking on other links. This action would be considered to have a ‘positive relevancy’ rating, because the length of time on page would suggest they engaged with the content, and they want to look at further sites to find supplementary information.
3. Another setting is that a user clicks through to your site and returns to the SERPS to refine their query before hitting ‘enter’ and searching again. This has neither a positive nor negative reflection on your site, instead indicating that the user recognises their search wasn’t specific enough to return the information they were looking for.
4. The fourth and final example is where a user clicks through to your site and returns to the SERPs to change their search query entirely. This action is also considered to have a ‘positive relevancy’, as the user is thought to have been provided with the information they required and has moved on to their next, unrelated search.

I predict that, as the BERT update gathers speed, these metrics will begin to hold more weight; they will indicate to Google the pages and websites that are most aligned with the context and intent of a search.

How can you improve on-site behaviour metrics?

Honestly, spending time researching what your audience is looking for provides good impact to your strategy. There are plenty of tools that help you do this, AHRefs, SEMrush, or even Google Search Console.

Consider the fact that there will be different levels to queries depending on where they are in the buyer journey. The information might need to be presented in different formats or geared towards varying queries.

Another (often untapped) resource is the sales or customer service team. What questions are they often asked by customers? Are there any barriers they are regularly presented with? It makes sense for these departments to be aligned and regularly collaborate.

I recommend using long-tail keywords and ‘questions’ found in tools such AHRefs to guide H1 and H2 meta titles, or structuring blogs as ‘FAQ’ pages, to provide concise, relevant and comprehensive information for each query.

I also recommend working on a writing style and tone that resonates with your audience and encourages users to engage with the copy. This makes them stay on the page for longer to read the content in its entirety. You may want to carry out some testing using a variety of styles on landing pages and monitoring the time on page and conversion rate in Google Analytics to discover some best practices and top-performing content.

It’s important to note that while long-form content DOES perform well, word count is not an indication of relevance when it comes to the SERPS under the BERT algorithm update.