One of the things we do here, aside from trying to help others to figure out how SEO and Google works, is to deeply research any pattern and data we can, in order to best figure it out for ourselves. We keep a close eye on Google and the likes, ask questions and sometimes even make ranking experiments to make sense of what goes on behind the scenes. But as is the case with all human observation (from quantum physics to sociology), it is impossible to study something without – at least at some point – becoming part of what you study.
1. The Story – 20k Visitors from Reddit = Improved SEO Rankings
We became part of our own study recently, when an article we published 4 months ago got picked up by Reddit, the social media network or self-professed “front page of the Internet”, in one of their sections (called subreddits). The article was a prediction attempt at how Google might “read” and rank your images in the near future and it was adequately picked up by the Futurology Subreddit. Whether Reddit really is the front page of the Internet is open to debate but the community is definitely a force to be reckoned with: the futurology subreddit alone boasts (as of now) with over 2.2 million subscribers. Not all of them may have been actively accessing the subreddit at the time our article got onto it, but we definitely felt an immediate and significant effect of our presence over there. What is interesting, beyond the effect itself, is how to interpret it, as we will shortly discuss.
2. The Huge Referral Traffic Spike on cognitiveSEO
The immediate effect that we saw on closer analysis was the impressive (if you look at the chart below) spike in site traffic around the day it was picked up by Reddit (January 5th). Within 1 day, the traffic on our blog hit 20 000 visits, an impressive feat by any standards, but also a quite singular one in recent history if you look at the rest of the chart. Of course, quantity does not necessarily equal quality, as evidenced by the no-less-dramatic drop in average session duration (almost inversely proportional to the spike in the number of visitors). On average, the users that day spent about 9 seconds on our pages. But before we sink into the pool of “youth-these-days” disappointment, remember that averages are statistical tools which can easily deceive. In this particular case, the average is a relative measurement which sometimes blurs the facts: it is likely that the absolute number of people who spent a really long amount of time on our page was still as high as (if not higher than) on any other day. Their numbers simply got cancelled out in the sea of users who clicked and bailed in just a couple of seconds.
Statistical clarifications aside, the fact remains: dramatic site traffic increases in the number of visitors does not necessarily mean a dramatic increase of the quality visitors.
3. The Increased Traffic – Statistics for the cognitiveSEO Blog
So what exactly happened to so many users that felt the need to leave the page as soon as they got there? Because the bounce rate was at an all-time high (96.19%), meaning that a vast majority just clicked, glanced and left in a hurry, just like you can see in the chart below.
Accidental clicks (followed, naturally, by hurried departures) are probably responsible for a large portion of the bounces, but surely they can’t all have been accidents? The profile of the reddit user is also something to take into account at this point. They are likely avid readers, parsing through articles in rapid succession without necessarily scanning the title in-depth before clicking on it and only deciding on whether they want to read it or not just seconds after the page has loaded. Another scenario to consider is that Reddit is first and foremost a community and its users are primarily preoccupied with interaction (even when it’s for the purpose of finding out information). Comments section are a lot of times the actual prized content and not just a redundant annex to the content. It is then not unthinkable that while most people who would comment on the topic would want to read the article, there was at least a considerable number of those who were more interested in the comments than in the article itself. There are probably a lot of other factors at work here to explain the staggering bounce rate. Despite all the bounces, however, the mere presence of so many visitors in so little time led to an even more impressive effect.
As expected and presented above, the average session duration was very low. Yet, as the increased traffic went away, the average session duration increased. This means that our usual readers platform maintained their behavior and after all the buzz in our site’s traffic, things got back to normal.
4. The Keyword Ranking Improvement on Ultra Competitive Keywords
Because the article was all about Google images and talked about this a lot, it actually got noticed by someone else, namely by Google itself. Well, it was probably noticed the first time around, but this time it got really noticed. Like that time in college someone threw a party and everyone came. People knew that person before, but after the party they really knew him/her. At least for a while. Much like this fictional American themed example, our fame was intense but short. We had a truly impressive search rankings spike for what is undoubtedly a very competitive phrase (“google images”), getting from the 74th position to the 8th in only one day! That is a feat by any standards. Yet, as you can see from the chart below, after just a few days our website rankings had a big downtrend.
We won a lot of points in terms of SEO ranks also, going from around 500 to around 1350 (almost three times more). Alas, our visibility fell less abruptly than it rose, but by the end of the month the levels had gotten back to where we started. By then, it was probably obvious to Google that what had happened on the 5th had been an isolate case and not the rule.
Although the saying goes “impressions don’t buy, real people do, we cannot ignore the high number of impressions registered on the 5th of January: almost 60.000. For a quick reminder, impressions are the number of times a post is displayed, whether the post is clicked or not. People may see multiple impressions of the same post as, for example, someone might see a page update in news feed once, can see it a second time when a friend shares it and so on. Even so, as you can see from the chart in the Google Webmaster Tools , the number of impressions reached for the queries containing “google images” is impressive and need to be taken into consideration as an important factor in our present research.
5. Did The Huge Amount of Referral Traffic Affected Our Google Rankings? – Data Correlation
Our 15 minutes of fame lasted a bit more, but not by a lot. Everything happened on the 5th, as the organic traffic spike confirms. You can see this yourself, in the screenshot below, taken from Google Analytics.
As more and more people saw the thread, up-voted it and commented on it, the post got pushed higher and higher and started to attract even more views and comments. Soon enough, it was all a matter of the rich getting richer. As with most threads on reddit, the buzz eventually died out and other posts took turns at the popularity contest. Posts usually lose popularity on reddit because others start to down-vote them and balance out the up-votes. This usually happens naturally, but sometimes it can also be influenced by users with administrative rights so as to keep a steady influx of new threads. Either way you look at it, the decrease of the organic traffic was a normal occurrence. The interesting fact was the direct influence the increase (and decrease) of the number of visitors had on the increase (and decrease) of the page and SEO ranking in the search engine. An interesting article written 4 years ago brought into discussion the same issue, still topical apparently:
Can traffic increases influence the keyword ranking positions in the SERPs ? It looks like it could be.
Yet, the author also stresses on the importance of concentrating more on obtaining links from webpages that can convert traffic rather than looking at the outdated metric of PageRank for example. So, concentrating on obtaining quality traffic, from social media or not, is crucial and you can monitor that traffic using free tools Google Analytics. Nevertheless, we cannot ignore the benefit of referral traffic in terms of how it impact SEO rankings within Google. As in our case, it seems like obtaining links that provide a high volume of referrals and click through data impact search rankings within search engines. We cannot draw a 100% certified conclusion here but we can assume, having our own experience as a starting point that referral traffic may influence search rankings. More research needs to be done on this matter in the future (and we are probably going to make an analysis of this kind in the near future) but I think that our findings on this situation is one that needs to be exploited and one from which there are a lot of lesson to be learnt.
6. Could the Inbound Links Rank Signal Boost Our Google Rankings Actually?
Of course, this whole analysis is basically trying to establish a cause-effect relationship between two elements : the referral traffic and the SEO rankings. The basic conditions seem to be meet: one happened before the other, there was a co-variation in one that corresponded to a co-variation in the other and the two are relevant to each other. However, we have ignored one additional criterion of a causal relationship: that there are no alternative causes that could lead to the same effect.
Also, I don’t think that the brand awareness generated via the Reddit link could have led to these increased SEO rankings. The moment the organic ranking improvement happened correlates directly to the time of Reddit posting.
There is one such possible alternative in our scenario: the Reddit link that was posted was a dofollow link that could have potentially influenced the ranking of the page. That inbound link merged into a broken link after a day, as it went further down the Reddit Category as new Reddit posts were added. Looking at our analysis below, you can see that the subreddit has both a high link authority and a high domain authority. While the actual measurable value and impact of the referral link is not very clear, it is definitely a different factor to take into account than the simple increase in number of visitors.
Could only one referring domain cause such a spike in rankings and in only a couple of hours since it was published?
We cannot be extremely sure about this but is not unlikely though.
We picked up some of the Reddit Links and analyzed them in order to see if the scenario where one domain could have boosted up so high was possible. As we browsed around the links we observed that indeed, most of them were links with good domain authority, yet, nofollow.
7. Does an Increase in Traffic Affects SEO and Google Rankings? – Conclusion
So could one inbound link from a high authority domain improve the search rankings and organic traffic to the top almost instantly for such a competitive keyword or did the actual traffic signal boost it? This is open to testing but my gut feeling tells me that there is a combination of the two signals and that both links and traffic contribute to such considerable spikes like the ones we have seen in our case.
As another personal experience, I noticed that social media traffic boosted some of our other posts when people started to share them more aggressively.
There are lot of things to discuss here and as listening to others’ opinions and analyze them is one of a researcher’s best skill, we would be more than happy if you shared your opinion on this matter. So, what do you think?
Does referral traffic influence SEO rankings? Or can only one referring domain boost a site’s rankings from a low position to a top position in only a couple of hours?
Is external traffic a Google ranking factor? Could this be an SEO technique that can be scaled?
Do you have any similar stories?