TweetSerping – You heard it here first
The following post is a recap on the presentation we recently made at the Brighton SEO conference on the subject of what we have nicknamed, Tweetserping.
Not only have we included the presentation notes below so you can check and try this for yourself, we have also included all of the slides used at the conference, we are just that kind!
TweetSerping. You heard it here first.
This isn’t entirely new and although there have been some tests carried out, notably by the good people at SEOmoz, they haven’t always been conclusive or specific enough to effect any changes in the way we conduct SEO. What we really wanted to know is not only if Tweets affected ranking positions, but how and could they help get a page that was previously impossible to index by traditional means, indexed faster.
The first test we ran was to demonstrate how closely aligned SEO and Social Media can now be. The piece itself was reasonably well written, but a little ranty (on purpose) and the title was designed to be as linkbait friendly (oxymoron?) as possible. We entitled the piece (which was pro-social media) “There are NO social media experts” with the term we wanted to rank for being “Social Media Experts”.
The key points to remember on this test are:
1. We placed the piece on our own blog (relatively powerful, good backlinks and PR4)
2. The blog fed in to Google News
3. Internal linking on the site was strong
4. The content was relevant to the site and blog
5. Fully optimised meta tags
We published the piece at 9.10am and by 9.15 it had not only been indexed, but appeared as the top story in the news section for the term “social media experts”. We also ranked at 395th within 5 minutes and just 5 tweets. An hour later, we had more than 20 tweets and ranked 146. Within 10 hours, we had climbed to the first page with over 100+ tweets and by the following morning, we ranked at number 3 for social media experts. We also ranked at number 2 for the singular version.
The tweets were initially put out by an account with more than 1,500 followers and in total, the piece was tweeted or RT’d by more than 130 different accounts. One of the most powerful tweets came from a user with nearly 5,000 followers. In total, the tweet reached more than 53,000 followers of those who tweeted it.
The test partially proved our suspicions, that tweets do, despite what Matt Cutts recently said, play a large role in helping a page rank for specific terms. Whether they do it alone however, is still unclear. What we can say is that tweets undoubtedly helped and played a role, but possibly at a secondary level considering how powerful the original blog is already, Google news helping the piece index very quickly and the relevancy of the content to the site it was posted on. It was clear, we needed another test.
With the second test, we eliminated most of the benefits found in test one.
The key points for Test Two were:
1. We placed the article on a website with little power and no internal links pointing to the article
2. The blog did not feed in to Google News
3. Internal linking was non existent
4. The content was completely unrelated to the website
5. We left the page lingering in the website for 4 days and it did not get indexed by Google, despite the bot crawling the site.
6. We optimised the title tag only
We published this piece on Thursday and by Monday, the page had still not been indexed.
In this instance, we used the same approach and I began with a tweet which would try to rank for the term “SEO vs Social Media” on Monday morning at 10am. Unlike the previous test, which got picked up and indexed within five minutes of its release, it took almost 2 hours 30 minutes before Google picked up. You have to remember however, the page had not been indexed for more than 4 days until we began tweeting 150 minutes earlier.
Upon checking the page at 12.30pm when it got indexed, it still had NO back links coming in, so there was a high probability that Tweets had played an almost singular role in getting the page indexed.
At 3.30pm after further tweets and retweets, the page ranked at number 60 for the term “SEO vs Social Media”. 2 days after the first set of tweets, the page still had no back links and now ranked 29th.
In terms of the accounts tweeting this message, only 2 had more than 1,500 followers; however one account alone had more than 12,000. Combined, the follower mass accounted for around 18,000 users via 36 tweets.
One further point here is that occasionally back links did show up (but never more than 3), which quickly disappeared on day 2. When back links did show up, they were daily tweet newspaper style pages.
Whilst it is still not entirely clear that tweets alone can get a page indexed or help climb rankings, what is clear is that tweeting out a page on relatively powerful twitter accounts DOES help SEO. It is entirely clear that without having tweeted the second piece it may not have been indexed for many days. We also placed a similar page on an equally low key website in the same circumstances and 10 days on, without a tweet, it still hasn’t been indexed.
Additionally, it is worth noting that whilst tweeting undoubtedly helped get the page indexed and helped improve ranking position, these tweets alone may not be the sole reason for the ranking position improvements. Other websites, even paper.li daily story sites that crawl twitter for specific keywords definitely play their part (even for a day or two) as do browsers & toolbars that monitor usage and report back to their respective search companies.
There is also no doubt that having a blog in Google News, combined with tweeting is a powerful way of leveraging early indexation and ranking gains, particularly with fresh content.
In fact, I’d go as far as to say that freshness of the content is the key to both tests and we shouldn’t ignore the power of good quality, regularly updated content, not just on blogs but on deeper pages too.
The next step
The final test is to rank better for a page which is already indexed and placing well on a competitive term where the content on the page doesn’t or hasn’t changed and this is where you guys come in :) Give it a go and let us know how you get on.
We also need to consider longevity of the positions gained. It has been too short a time to measure this in any depth, but we’ll continue to monitor the two terms we have worked on and see if they rise, fall or maintain and in turn report back through the Fresh Egg blog.
- Make sure blogs are in Google news
- Build quality followers on Twitter (above 1,500 seems to be helpful)
- Tweet your posts using keywords you want to rank for
- Optimise at least the titles and make sure the titles match the Tweet
- Try and encourage other powerful accounts to RT
- Paper.li and daily newspaper style websites can help (temporarily)
- Fresh content in page copy helps the likelihood of ranking improvement (still needs further testing)
You can download the slides that go with this presentation HERE.