The professional community has been discussing the evaluation of search engines based on social signals for a long time. We often see skepticism in this area, so I will add my own data and experiments to the studies made by reputable agencies. Each experiment in SEO has complex input conditions, so its results are not always absolutely verifiable. That is why my experiment to confirm social signals has quite clear input settings to eliminate other factors.
What social signals are and why search engines need them
Search engines want satisfied users. And users want quality search results. At present, SERP results can still be manipulated. Give me a large enough budget for a sufficiently large professional SEO team, and I will put any website onto the first page of a search engine in a few months. Google is trying to fight this primarily by purifying its SERP with Panda and Penguin updates.
Social signals could be the desired solution. Spontaneous sharing of URLs, mentions and social interaction is difficult to manipulate and it could represent the quality and popularity of a URL from a visitor’s perspective more accurately.
Unfortunately, search engines, as well as other signals, are vague; hence, experiments like this one to confirm the assumptions about the power of social signals.
Hypothesis: When are social signals influential?
To make social signals work, search engines have to index the social media. The social media have never helped engines much in terms of website throughput. Frequent use of Ajax forced Google to enhance its crawler; so, theoretically, there are no technical obstacles to prevent indexation.
Experimental conditions
The purpose of my experiment was to determine if search engines can index URLs linked only from Facebook, Twitter or Google+ without other links from any other websites. That is why I created the following pages:
Each page had a unique content and each was linked only from the social media it targeted. The experiment began on June 1st 2013. I created a simple post with a link on each social network (marked as nofollow on Facebook and Twitter and shortened via bit.ly on Twitter) open to all users (without any restriction of user groups or a locked status).
|
Social network |
Number of shares |
Number of comments |
Number of Likes/Favorites/+1s |
|
22 |
25 |
6 |
|
|
19 |
7 |
0 |
|
|
31 |
4 |
9 |
Up to July 4th 2013, I checked if the mentioned URLs were linked using a report from the Majestic SEO fresh index:
Unfortunately, even this tool could not follow all the factors affecting my experiment, so I have to admit that the various aggregators of Google+ and Twitter posts undermined the experimental conditions and the following URL links leaked out into the “live” Internet. The leaked links include:
http://tweettunnel.com/Berkofff (Twitter)
http://new.tweettunnel.com/reverse2.php?textfield=radomirpa (Twitter)
http://twitter.mracik.sk/account/hanickapavlin (Twitter)
http://it.circlecount.com/cz/explore/ (Google+)
http://www.mygoogleplus.com/timeline.php?id=106656366093187516063 (Google+)
http://cpedialog520.appspot.com/profiles/106656366093187516063 (Google+)
Therefore, we do not know if the experimental results are verifiable for Twitter and Google+. Paradoxically, the only experiment that did not leak (according to the search engine index) was the experiment on Facebook. Based on the results, we can make specific conclusions. What were the experiment results?
|
URL |
Date of indexation – Seznam.cz |
Date of indexation – Google.cz |
|
http://seojedobro.eu/facebook/ |
||
|
http://seojedobro.eu/twitter/ |
||
|
http://seojedobro.eu/google-plus/ |
Google managed to index all the mentioned websites; Seznam managed to index all websites except Twitter.
The importance of pages
An interesting aspect is the ranking of the tested URLs in the indexed page listing from seojedobro.eu via the site: operator. In both of the tested search engines, we can see that the URLs have a relatively high ranking in comparison to other domain pages (which are fully optimized – they are also linked from other internal and external links.)
|
URL |
||
|
http://seojedobro.eu/facebook/ |
44. |
2. |
|
http://seojedobro.eu/twitter/ |
– |
3. |
|
http://seojedobro.eu/google-plus/ |
3. |
4. |
The results of the experiment
It was verified that search engines have no problem with indexing social networks (including Seznam.cz). We also know that other than Twitter, with which Seznam.cz had problems, it is enough to link from a social network to force a search engine to browse the URL. That is why we know that the pages on social networks can be evaluated like any other web pages and the links can be regarded as common (in some cases nofollow) links.
We also know that the publication of links on Google+ and Twitter can generate more links in aggregators.
It was also confirmed that pages with significant social interaction are positively evaluated in search results (see the ranking using the site: operator).
This was only a simple experiment to verify the effect of indexation from social networks. Of course, one could go much further and test the effect of various types of content on social networks or which comments, shares or “Likes” are most effective. There are options.
What does this mean for online marketing?
Social signals affect website ranking. Therefore:
- Each post on a social network strengthens the website you are writing about or linking to – and not just in terms of SEO.
- Use the right keywords when writing your post.
- Do not forget to optimize the on-page factors of your profile on social networks.
- Increase the involvement with appropriate text (a task for a copywriter), timing and topic (using keyword analysis).
See you again by more experiments – and share your experience!
Pavel Ungr
This post was written by Pavel Ungr: follow him on Twitter, connect on LinkedIn or check out his Czech blog.
