Can social media help SEO? Before answering this question, here’s a quick glossary of common search engine optimization terms for readers who may not be SEO experts.
Glossary of SEO terms
- SERP: Search engine results page
- Search order: The position a URL holds in a SERP for a particular keyword
- Search visibility: A metric used to calculate how visible a website or page is in the SERP. For example, if the number is 100 percent, that means the URL is ranking in the first position for a keyword(s). Search visibility is especially important when tracking a website’s overall ranking for a keyword basket.
- Domain or page authority: The strength of a website or page in the eyes of search engines on a particular subject. For example, the Moyens I/O blog is perceived by search engines as an authority in social media marketing. This means we have a better chance of ranking for social media related keywords than a food blog like Smitten Kitchen.
Does social media help SEO?
The question of whether social media has any effect on SEO has been debated for a long time. In 2010, both Google and Bing admitted to using social signals to help rank pages in their results. Four years later, that attitude changed after Twitter temporarily blocked access to Google’s social networks. In 2014, Matt Cutts, Google’s former head of web spam, released a video explaining how Google cannot trust signals that might not be there tomorrow.
This is where the conversation stopped. Since 2014, Google has publicly denied that social media has a direct impact on rankings.
But now it’s 2018. A lot has changed in the past four years. One notable change is that social networks are starting to appear on search engines on a much larger scale.
Facebook URLs in the top 100 on Google.com (US)
Twitter URLs in the top 100 on Google.com (US)
Have you noticed that Facebook and Twitter pages are growing exponentially with Google’s results? We did well and thought it was time to analyze the relationship between SEO and social media with a series of tests.
Say hello to “Project Elephant,” an experiment named after the “elephant in the room.” The elephant in this case is the long asked but never answered question: Can social media help improve search rankings?
How we structured our experience
Representatives from Moyens I/O’s inbound marketing, data analytics and social marketing teams came together to develop a reliable and controlled testing approach.
We organized our content (blog articles for the purposes of this experiment) into three groups:
- Control group: 30 articles with no organic posts or paid promotions on social media (or elsewhere)
- Group A (organic only): 30 articles published organically on Twitter
- Group B (paid promotion): 30 articles posted organically on Twitter, then featured for two days on a $100 budget each
To simplify the number of data points, we chose to run this first test on Twitter and created a release schedule to keep ourselves on the right track.
But before we started the test, we needed to level the playing field. That’s why, exactly one week before launch, none of the 90 articles selected for trial were updated or promoted. This allowed us to establish a baseline for search rankings.
Following this step, we promoted two posts per day from Group A and Group B over a two-week period and measured results over the next week. From start to finish, the entire experiment took about a month to run.
methodology
To ensure we cover all our bases, we recorded the following data points:
- What keywords were we tracking?
- What URLs (blog articles) were we tracking
- Monthly search volume for each keyword
- Google search ranking of each article before testing started
- Google search ranking of each article in 48 hours after that testing started
- A weekly Google search ranking of each article after that testing started
- Number of links pointing to each article before testing started (backlinks are the number one driver of search rankings)
- Number of unique websites pointing to each article before testing started
- URL rating for each article (aHrefs metric, more in a minute) before testing started
- Number of links pointing to each article after that test concluded
- Number of unique websites pointing to each article after that test concluded
- URL rating for each article (aHrefs metric) after that finalize the test
As we walked in, we realized that the accepted position on the matter was: there is an indirect relationship between social media and SEO. That is, content that performs well on social media will likely gain more backlinks, which will help boost search rankings.
Because of this indirect relationship between social and search rankings, we needed to be able to explain whether traditional domain/page authority metrics played a role in any ranking changes.
Page authority metrics were based on aHrefs’ Live Index. aHrefs is an SEO platform that crawls web pages and collects data about the relationship between websites. They’ve scanned 12 trillion connections to date. aHrefs’ web crawling speed is second only to Google.
Experiment results
From a high level, we can observe an improvement in search visibility across the three keyword baskets. As you can see from the above results, strong relationship between social activity and rankings.
Let’s focus on the actual data points to better understand the mechanisms behind the increase in ranking.
As shown, the control group sees the lowest levels of rank improvement and the highest rank reductions compared to the other test groups.
While the rankings were saved during testing, we specifically wanted to reset any changes that occur right after a promoted content on social media.
The scatterplots above show the change in ranking and the total number of social interactions observed in the first 48 hours of a shared content. As you can see, the organic and amplified test groups outperform the control group, where more ranking losses were observed.
The table above specifically looks at the change in rank over the first 48 hours versus the total number of social interactions associated with this content asset across all test groups. Looking at the surface data, we can observe a positive linear trend line showing a positive relationship between the number of social participation and change in ranking.
Of course, any seasoned SEO strategist will question this correlation because of a number of factors related to how social interactions can affect other metrics that are actually ranking factors. Later.
Looking at the total number of social interactions versus change in rank in all test groups after one week, we can also observe a positive linear trendline indicating a positive relationship between the two measures.
What about social activity leads to more links, which in turn leads to better rankings.
As mentioned above, Google has traditionally refuted the fact that social activity affects ranking, instead arguing that social engagement can affect other metrics that can affect your ranking, such as links. This chart shows the change in referral domains pointing to a promoted content and the number of social interactions it receives. As we can see, there is definitely a positive correlation between the two measurements.
SEO experts can keep scrolling because they already know the answer to the question of whether links are associated with better rankings. But social marketers should listen. The charts above show the ranking against the number of referrer domains that indicate content presence in context.
As you can see, there is a strong correlation between the number of websites pointing to a piece of content and relative ranking. For fun, we filtered the results by search volume and observed a much less significant correlation for keywords with more than 1000 monthly searches, indicating a higher level of competition. This makes sense. You will see much greater improvements in less competitive conditions for every link earned compared to more competitive conditions.
What if we remove the instances where we observed a change in the driving domains?
To properly challenge the theory that social marketing can only affect rankings through acquired links and not directly, we removed all examples of keywords that observed a change in leading domains during the testing period. We were left with only two factors: rank change and social interactions.
Admittedly, this level of filtering emptied our sample size, but left us with a promising picture.
There is a positive relationship between social participation and rank change.. Overall, there was greater improvement in rank associated with social engagements than observed rank losses.
Of course, this data encourages a larger-scale test that would be difficult to extract given the strict SEO and social methodologies applied to this experiment.
What marketers should (and shouldn’t) do with this data
Yes, it can help with social SEO. But that shouldn’t give you a free pass for overposting and spamming people’s posts. If you do, you risk annoying followers. And then they may ignore your posts, or worse, unfollow you altogether.
The quality of the posts – not the quantity – is key. Yes, regular sharing is important, but it’s pointless if you’re not delivering value to your audience.
Remember, a URL may only need one new backlink to significantly improve its search rankings (depending on how competitive the keyword is and how authoritative the site is to your link). If you attract enough of the right people to share your content on the website, you will see an increase in search rankings and search visibility.
Social marketers should also consider the effects of paid promotion on SEO. Indeed, our findings show that paid promotion almost doubled the SEO advantage of organic promotion.
SEO should be thoughtfully integrated into your broader social marketing strategy, but it shouldn’t be a driving force. if you focus on creating and sharing quality content, you will be in good shape. After all, quality is the number one ranking factor on Google.
Use Moyens I/O to share quality content across all your social media channels from a single dashboard. Grow your brand, engage with customers, keep up with your competitors and measure results. Try it for free today.
To start
