Obviously getting higher organic search rankings on Google is critical. But figuring out the idiosyncrasies behind how Google computes their rankings is an incredibly difficult task!

This month, SEMrush conducted extensive research into search ranking factors via a novel machine-learning-based analysis of 600,000 keywords, pinpointing 17 elements that appear to impact the position of sites and search listings in the search results. And, by using their insights, you can make changes designed to help improve the rankings of your domain and pages:

 inline image

Image source: SEMrush

Here's an overview of their findings.

Direct Website Traffic is the Most Influential Ranking Factor!

This was an extremely unusual finding because generally "direct website traffic" (when a user heads straight to a site rather than say, navigating via some other channel like a search engine results page) isn't considered a traditional "SEO Ranking Factor" that Google would measure and use, therefore many SEOs aggressively disputed this particular finding.

Rather than re-hashing all of the arguments about the study, I'd rather focus on trying to explain how this particular finding might have been possible.

I believe that that a higher share of "direct website traffic" is essentially just a strong indicator that the domain in question is considered to be more of an authority (or brand) in the niche.

If you're a bigger brand, you'll tends to have more links to your domain in the first place, which is obviously an important ranking factor.

The fact that direct website traffic was so strongly linked to remarkable SEO rankings, demonstrates to me that the future of SEO is increasingly more about conventional marketing activities, like working to build a strong brand image (via content marketing, advertising, etc.) in order to create the brand awareness and affinity among your target market, which in turn drives the unusually high CTRs which are rewarded with higher search rankings.

User Behavior Plays a Huge Role

Considered a sign of a page or domain's quality, three user behavior metrics fill out the top four most influential factors.

A visitor's "time on site" is a figure that represents the total amount of time they spent viewing pages on your domain during a session while "pages per session" is a count of the number of pages they viewed. Both of these metrics could be considered by Google as signs of quality and authority, as the user was engaged by the content and chose to spend more time on the site as a whole.

The "bounce rate" is a reflection of how many page visitors left the domain after viewing just a single page. Higher ranking sites almost always have lower bounce rates as Google may see higher bounce rates as an indication that the contents of the page weren't actually relevant to the visitor, creating questions regarding the quality of the domain.

Now, it should be noted that a person leaving the site after viewing one page doesn't necessarily mean that the content wasn't helpful, authoritative, or valuable. However, an extremely short session duration or unusually high bounce rate certainly suggests that the visitor did not find what they were looking for. 

My own research into this matter found that Google uses "dwell time" to re-order search results - it's essentially the same metric as all three of these aforementioned user engagement metrics, just measured from the Google SERP rather than the traffic receiving website.

Backlink Factors Tie for 5th

Four backlink factors, new additions to the research, actually tie for fifth place in regards to importance.

The study found that the total number of referring domains, total backlinks, total referring IPs, and total follow-backlinks all played a significant role in a domain's SERP position.

Keep in mind that "direct traffic", the number one "ranking factor" in this study is obviously correlated with having more links since bigger brands attract more links.

Google has previously confirmed the assertions found here, which show that have a robust backlink portfolio is vital if you want your domain to rank highly, especially on high-value keywords.

Each of the backlink factors are interconnected, so improvements to just one of these areas isn't guaranteed to yield positive results. Instead, you must view these as part of a package.

If your domain doesn't have a suitable backlink portfolio, then focus on link-building strategies to help improve your site's overall ranking. One method is to concentrate on low-volume SERPs, as there is less competition based on the keywords and it may be easier to garner attention for your content and score valuable backlinks.

My view on this matter is that link profiles still play an enormous role in filtering out the top hundred or so search results among potentially hundreds of millions of competing documents, but that the user engagement rates like CTR and Dwell Time play a decisive role in ordering of the top ten contenders for any given SERP.

As an analogy: having a strong link profile will get your page into the playoffs, and having strong user engagement metrics will win the championships. Or alternatively, links are necessary but not sufficient to win a competitive SERP.

HTTPS and Google's Goal for a Safer Web

Google has taken a hard stance about making the internet a more secure place, and this includes favoring HTTPS sites in their rankings. By transitioning or maintaining an HTTPS, Google views you are more of an authority and rewards your efforts to protect your visitors.

For high-volume keywords, HTTPS adoption is almost mandatory if you want to rank highly. For low-volume keywords, HTTPS is less critical, though it can give you a leg up in the competition by distinguishing your domain from the crowd.

Ultimately, we all understand the benefits of ranking highly in search engine results, and the report by SEMrush practically creates a guide to help you do just that.

Keywords in Page: The Donkey of SEO Ranking Factors

Rounding out the very bottom of the ranking factors are keywords in the body or headline, or keyword density has the lowest impact on search rankings. Why? because page titles loaded with keywords (eg: "Cloud Backup Storage - Best Cloud Based Storage Solutions") are ridiculous and yield low CTRs which yield terrible rankings. Instead of a stupid keyword-stuffed headline, try doing this instead.

Additionally, Google has confirmed that meta keywords, for example, are not used at all for ranking purposes.

Crazy SEO Experiments: Final Thoughts

Despite the industry criticism of the SEMRush study conclusions (particularly relating to "direct traffic" being labeled as a "ranking factor"), I still believe that there are still a few interesting data points in there that can discuss and learn from.

I've personally run dozens of my own SEO experiments over the past two years regarding the impact of user engagement factors like CTR, bounce rate, and time on site on search rankings, and have found them to be more or less consistent with the SEMrush study: That they have an enormous effect on how your domain and/or content ranks with Google's organic search algorithm.

Additionally, since click-through rates are so critical, so you need to make sure you write an intriguing headline to draw people in and then keep them engaged with your fantastic content. Yet, since visitor behavior is vital to your rankings, misleading clickbait strategies will lead to low time on page and high bounce rates, and will damage your rankings, even if your CTRs soar. Both high CTR and high dwell time are essential for strong SEO.

Finally, the one of the best ways to improving your rankings is to market the heck out of your brand. Even if you don't believe as I do, that brand affinity has a huge impact on links, CTR and search rankings, it will most certainly increase direct website traffic which is a worthwhile business goal in of itself, regardless of if it impacts rank.