Search engine optimization (SEO) is an industry that’s always changing, and the better plugged in you are to news of things like algorithm updates, technological changes, and new techniques, the further ahead of the competition you can get. Fortunately, the SEO community is a relatively open one, and everyone—from the dominating influencers to the one-man operations—is willing to share the latest information they’ve gathered on how search engines work.
One of the latest reports I’ve seen is from SEMRush, researching some of the latest ranking factors known to influence search engine rankings. I gave this report a thorough read-through, and walked away with several new insights about ranking factors in 2017.
A Note on Research Methods
It’s also important to note how this research was carried out, as not all SEO research is cut-and-dry. SEMRush has over 1.5 million users, and mined data to research 600,000 different keywords from all over the world. Researchers compared various pages’ positions in SERPs with a list of purported ranking factors, and used the Random Forest machine learning algorithm to track consistent patterns (rather than relying purely on correlational observation). Then, researchers narrowed the scope to the top 10 SERPs for each keyword, and double-checked the data for the 20th position entry to verify if their perceived patterns held true.
Then, researchers categorized data based on different “volume classes,” as they hypothesized that different competition niches may bear different results. They also attempted to categorize data by different levels of keyword competition, but results remained (more or less) the same. In some cases, there were major differences between head keywords and long-tail keywords (of four or more words).
Biggest Takeaways
These are some of the most important insights I gleaned from these data:
1. Website visits are huge.
We’ve always known that traffic plays a role in determining search rankings, at least to a degree, but SEMRush lists website traffic as the most important indicator for high-ranking positions. At this point, you might be wondering about the relationship between traffic and high rankings; of course higher-ranked sites will see higher rates of traffic, because they’re more visible than their competitors. It’s a classic chicken-and-egg scenario. However, there’s clear data to indicate that a change in traffic alone could be enough to cause a site to increase in rank. This represents evidence for the thought that click-through rate does affect search rankings. So, make sure to optimize your meta descriptions!
2. Inbound links will make or break your SEO campaign.
I’ve been an ardent supporter of the importance of link building since the early days of SEO, and few industry experts debate the fact that the quality and quantity of inbound links to your domain and individual pages are extremely important. Links are the primary gauge Google uses to evaluate authority and trustworthiness. What surprised me from this report, however, was the sheer number of links that separates high-ranking sites from low-ranking ones; the average difference between rank 1 and rank 10 was more than 10,000 referring domains. That’s a huge divide, especially considering how many marketers aren’t currently investing in a link building strategy, as I learned from conducting a recent study. My advice on link building remains the same as it’s been for years: if you aren’t already working on a link building campaign, you need to start. For help, see SEO Link Building: The Ultimate Step-by-Step Guide.
3. Keyword matching isn’t that important.
When Hummingbird was released, and in the years since, optimizers have debated just how important exact-match keywords really are. Some argue that Hummingbird’s semantic search capabilities make it so that keyword inclusion is more risky than beneficial, while opponents argue that keywords are still a necessary ingredient in a successful strategy. Now, the data from SEMRush’s research is in, and it shows that 18 percent of domains ranking for a high-volume keyword don’t contain that keyword anywhere in the body of their content. Long-tail keywords were, on average, even less present in body content than head keywords, so if you’re including more long-tail optimization to cater to voice searchers, you might consider focusing more on quality content than worrying about keywords. However, there is no evidence that including keywords hurts your efforts, so while it doesn’t seem to be important, my official recommendation is still to include them (at least, until we have evidence that it could be harmful to your efforts, which I think is highly unlikely to ever happen).
4. Content length is important.
Longer content isn’t necessarily better content, but it does tend to rank higher and attract more links and shares. This could be a result of several factors, including more content to crawl, more time spent on pages, or simply better-researched material. In any case, results in the top 3 positions tended to include about 45 percent more length than position 20 results. On average, that meant around 750 words on page, compared to 500 words on page for lower-ranking competitors. This is why I still like to aim for at least 1,000 words in my written content when I can.
5. User experience matters across multiple dimensions.
Ah yes, user experience—one of the most subjective, yet most important factors for ranking websites. If Google perceives your users aren’t enjoying your content, your site is going to rank lower. How can it tell? First, consider that time spent on site was found to be the second most important determinant of rank (next to overall website visits, which could be a secondary indicator of positive user experience thanks to return visitors). The third and fourth most important factors? Pages visited per session and bounce rate, further measurements of user experience. The average bounce rate for a top-3 site was 49 percent. I’ve made the case, in the past, that bounce rate and time-on-site can’t be used as ranking signal since it can only be calculated on websites that have Google Analytics installed, but I may have been mistaken. Google has other ways to get that data. For example, consider users who find a website in Google, visit it, then use the “back” button on their browser to return to Google. Google can tell how long they spent on the external website before returning to Google (and if they returned at all).
Advice and reports on SEO can be dubious at times, but I have to hand it to SEMRush for putting together an interesting and well-researched report. Ranking factors have changed significantly since the early days of SEO, and even though fundamentals like high-quality content and links still form the foundation of a good SEO campaign, the best tactics to achieve them have shifted tremendously.
I hope you’ve learned as much as I have about the shape of ranking factors in 2017, and I encourage you to revisit your strategies to accommodate that new knowledge.
