Update (Dec 10th at 10:30 AM PST) - Wow! The participation has been overwhelming! Thank you everyone! I am still collecting data and will be posting the full results in the near future. For those curious, early results actually show the tweet version winning.
Update (Dec 10th at 4:35 PM PST) - The Tweeted link is now ranking number #1 for "ending hunger sierra leone". Additional analysis coming!
Last week Rand Fishkin covered the new ways social media is impacting traditional SEO. While many SEOs had assumed this was happening already, this announcement marked the first time this new type of content endorsement (links within Facebook and Twitter) was actually confirmed as a ranking factor by search engines representatives. Although there has not yet been enough time to run a formal experiment on the impact of these type of content endorsements, I would like to get the ball rolling with an informal experiment that will both help start to answer the questions these changes have brought and at the same time promote the holiday spirit.
The Experiment
Disclaimer: This experiment is not like the more formal tests we run here at SEOmoz. Any results we glean from this test will only be hints at what is going on. Admittedly, there are many factors that this test simply does not take into account. Seriously guys, this is more a fun test rather than an actual scientific experiment. :-)
In order to test these two ranking factors (traditional links and social media citations), we are going to need to promote two nearly identical pages using one promotion method for one and the other promotional method for the other while targeting the same keyword phrase. We can then see which page performs better in both the short term (my guess is the the tweet driven page) and the long term (my guess is the link driven page).
These two test pages will need to be targeting the same keyword phrase and on the same domain. I could do this on my own domain but doing so would be disingenuous (Hey everyone, send links and tweets my way! It is for an "experiment" ;-p You wouldn't be fooled for a second and my apartment would get TPed... again)
Instead, the two test pages will be on a local charity's website and target the phrase "Hunger in Sierra Leone". The local charity, SeeYourImpact, is a registered 501(c)(3) nonprofit organization that sends 100% of donations to the charity of a donation makers choice. The non-profit then photographs the results and sends the photo and a story of how the donation made an impact on an actual person to the donation maker so they can 'see the impact' (thus the name) of their donation. There is absolutely no financial affiliation with SEOmoz and SeeYourImpact although we do do consulting for them on a pro bono basis. In short, we think they are doing a great service but no money is being exchanged between the two organizations.
We Need Your Help!
Below is the code for either tweeting or linking to one of the two pages on SeeYourImpact's website. If you are so inclined, please feel free to link and/or tweet about these pages. The pages are about trying to help eliminate hunger in Sierra Leone.
Tweet Option:
Help end hunger in Sierra Leone and meet the life you change https://bit.ly/hpsMJq
Linking Option:
<p>Help end <a href="https://seeyourimpact.org/explore/sector/hunger/fighting-hunger-sierra-leone/">Hunger in Sierra Leone</a> and meet the life you change.</p>
Help end Hunger in Sierra Leone and meet the life you change.
If tweeting or linking is not your thing (wait, what are you doing reading an SEO blog?), feel free to browse SeeYourImpact's website and consider making a donation.
Sharing the Results
I will be monitoring the results and sharing my findings in this blog post after they come in. Thanks everyone!
Tweeted.
My opinion is that the tweets will fast create a buzz for the website, which will make people to create links to this page.So tweeting will bring page links, interesting how many of the standart links will bring tweets about this page?
Do you plan to change the snippet or anchor text at all through the experiment? Just curious how having all links with same anchor text may or may not affect how this turns out.
- Tweet tweet -
Done via tweet - curious which option will win ...
HI Danny,
Nice experiment. I done the tweet but my guess is that ppl will not use the second option because is harder to do. IMO, is easier someone tweet something than create a link into his/her own site.
But... let's see what happens =)
Agreed! It is much harder to link than to tweet which is part of the point. Luckily, there have already been some traditional links added.
Would be interesting to see how much impact the Tweets have in comparison to the actual links, especially given the point you made - it's much easier for people to Tweet a link than it is to post it somewhere relevant on their website/blog.
Harder...schmarder! Maybe a bit tricky for "normal" people but I'd hope the crowd here can manage an actual static hyperlink. Sounds like a worthy cause, so that's what I went for. I did tweet the page with the link on it (as an afterthought).
Look forward to seeing the results. Look forward to more stuff like this on SEOMoz too.
Added a link.
Looking forward to the results!
Tweeted for you.
I, like yourself, believe that the Twitter option will have the best short term affects. I still agree with everyone that a traditional backlinking strategy is the most important for long term goals.
Looking forward to the results.
Tweeted. Can't wait for the results!
Looking forward to see the test results. I shared the link on twitter
"Last week Rand Fishkin covered the new ways social media is impacting traditional SEO. While many SEOs had assumed this was happening already, this announcement marked the first time this new type of content endorsement (links within Facebook and Twitter) was actually confirmed as a ranking factor by search engines representatives."
Danny, that makes it sound as though Rand's coverage was the first time the SM-to-SERP relationship was confirmed, but in fact Rand was writing about a post from Search Engine Land's Danny Sullivan, so it's really he who should get credit for reporting the information. His post is here: https://searchengineland.com/what-social-signals-do-google-bing-really-count-55389 .
Oh...and yes, I just Tweeted "Help end hunger in Sierra Leone and meet the life you change https://bit.ly/hpsMJq". Cheers!
Most of us regulars on SEO circuit know Google & Bing acknowledged this fact recently - but very few although knew didn't dare to say it simply because of something mesmerizing and accepted algo called "relative=nofollow".
And twitter/FB hardly give dofolow links.
So now the topic of nofollow is in serious .................
I want the social to win even though, having recently been to a SEO/marketing seminar here in Italy which seems to disprove this...
Added a link to my blog chrisfaron.com & tweeted at twitter.com/chrisfaron
Added link on my homepage.
Tweet gone - looking forward to the results :)
Tweet
I have added a link on our website, looking forward to seeing the results!
Done both ways.
As you I believe that Twitter will give faster results, even though - IMHO - for long run link building campaign the traditional backlinks are still more important.
Anyway, Danny, why is not there an option to share the link in Facebook too?
Good question about Facebook. When I run these search engine tests I prefer to only have two major variables. These tests are already messy and I try to keep them as simple as possible :-)
Will you be running a Facebook version in the future or does it depend on the outcome of this Twitter one?
Thanks
- Jenni
Tweeted - very interested to see the results and thanks for running the test
Just tweeted the link.. look forward to the results!
Tweet baby Tweet
Tweeted as well :D Great thing.
I'm curious :)Here's my contributions....
I already see the tweeted page showing up # 7 on Google. First page on Bing also, along with this page.
Are these brand new pages? Or is that to be assumed?
Hi, I added a link to my "Seasons of Life" blog at https://seasonsoflifeeternallessons.wordpress.com/2010/12/10/whats-your-favorite-charity/ and sent a tweet as well. Hope this charity does well in doing good, and I look forward to the results of the test.
I wonder if Google thinks this is suspicious... They are nowhere to be seen on Google.
On Bing, I see them at #1 and sometimes at #2.
Curious if Danny, or anyone else, knows where they ranked before the experiment started. If so, please share.
Tweeted, I look forward to seeing the results
I've tweeted - keen to see how this one goes...
I ran a similar test case in March of this year (I paid a small amount for health-related tweets on ad.ly for the experiment). My findings were inconclusive - there was no obvious pattern to ranking movements.
* Edit:
Here were my findings...
After test 2:
Across 7 keywords (synonyms) and 8 'search engines' (.com and the AU regional variations), we had 56 chances for ranking changes, and in actuality only one keyword instance rose from Twitter/bitly links, which was Google.com. The rest stayed the same or dropped slightly.
After test 3:
Once again, no consistency in the changes seen in rankings. Rankings in Google Australia stayed pretty static, or even dropped slightly. Yahoo saw the biggest improvement, with one keyword coming in at 12th place from being outside the top 50 twelve days prior. Bing was also static.
I tweeted it. It will be interesting to see the results.
So I am seeing the tweeted page #2 in bing and the linked page nowhere (at least not in first 5 pages or so) however as for google I don't see either page showing up. I noticed some people above saying they saw pages ranked so are those same people still seeing results or was it possible personalized results they witnessed?
I'm wondering if in fact it wasn't a personalization issue with google and there were results which are now gone if it is possible we have triggered some sort of filter due to a weird link, tweet pattern that looked unnatural. I pretty much assume the linked page would take a little longer to rank as it would take time for the crawlers to find all the backlinks and factor them, but I really expected to see the tweeted page to be showing up by now if in fact they are taking it into effect. It would make sense that the tweeted page would at least bring it up to the top on a temp basis because of it's immediate relevancy, so shouldn't we have seen something by now.
Is anyone else seeing google results? Thoughts?
Tweeted earlier and gonna provide a couple links to it later this evening. Should be interesting, but more importantly its for a good cause.
Great experiment, loving it. Here's my observations from Australia so far:
Tweeted Page = No. 6 in both G.com.au & G.com
Tweeted Page = No. 4 in Bing.com
Linked Page = Nada
Tweet sent! This is a pretty amazing test. I'm looking forward to the outcome.
tweet and wait for the result....
Tweeted. Interesting test, thank you for the continuing work and research. Well worth the Pro subscription.
Just from the comments I count 40 tweeters and only 6 linkers. I would say this ratio will get worse with retweets but at the same time I would think the Google has some idea of what are the relative values of each. I've linked, for the record :) Links will win once they are all crawled but I think we are overdoing it with the anchor text links, could do with some straight URL and brand links to balance it out.
Has any body tried searching with "fighting hunger sierra leone"? None appears to be top 10.
And the tweeted version ranked #1 because it used "ending hunger sierra leone",which is in the title and url of the page.
Does that mean the test is not yet finished?
I totally agree, I was going to post a comment about this as well, until I read your comment. This is not an accurate test because of the fact that one article reads, "Fighting for hunger in Serra Leone" and the other reads "Ending hunger in Serra Leone", so obviously the article without the word "Ending" in the H1 and title tag won't rank well for "Ending hunger in Serra Leone". If I search for "Fighting for hunger in Serra Leone" it outranks the other so their is no correlation to retweets = higher ranking.
This test is a false one and has put the SEO industry in a spin because of it...
To add to your comment, the search phrase "fighting for hunger in Serra Leone" retrieves the "fighting for hunger in Serra Leone" article in 4th position. The "ending hunger in Serra Leone" article is no where to be seen.
However, searching for "hunger in Serra Leone" retrieves "ending hunger in Serra Leone" article in 2nd position.
So maybe retweets do work... I'll need another test to be completely convinced though. The "fighting" vs "ending" is still a gray area for me.
How do you compensate for duplicate content? What were the standings before the test?
best,
Reg
nbs-seo.com
This should be interesting! looking forward to seeing the results. Great idea for a post also! its generated quite a bit of buzz! nice work
with a simple search I can see that some links to this page https://seeyourimpact.org/explore/sector/hunger/fighting-hunger-sierra-leone/ have already been indexed by Google but this page is not showing at all when searching for the targeted keyword. The problem could be that the 2 test pages have the same text and for this reason Google doesn't count one of these. What do you think?
Would love to see those results :)
It will be interesting to see the results.
Danny Dove no longer works in SEOMoz. We'll get no results :(
You can check this by yourself. Just open Google.com and...
Late to this post. About the linking option, isn't there a concern about all the anchor text being the same and triggering a filter?
Danny, how is the project tracking so far?
Interesting experiment. What is the ratio of tweets vs. links so far? It'd seem much more people would tweet it than actually link it due to the ease of tweeting. Looking forward to the results (specially when it can be measured short-term versus long-term).
Linked. Looking forward to the results!
-Evan
Just added a link and then tweeted the new blog post URL (gets Google informed of the page in about a minute). Created a blog post for it and it is the ONLY link on the site... I'm pretty interested to see how many Tweets were offered compared to Links, and the associated comparisons that will follow.
This is an awesome way to test this kind of thing Danny. I posted tweets on my two accounts as well as facebook, linked in and buzz.
hope they help :)
of course you will have to post the number of links vs number of tweets....
Done! via Twitter
Will add a link.
I'd be very curious to see what happens given equal amount of links to tweets, rather than tweet outnumbering the links.
But the real world will definitly see more people simply making the tweet due to its ease of use.
Thanks for posting this test. Social is heading to playing a bigger role.
I must say that overall I am really impressed with this blog.It is easy to see that you are passionate about your writing. If only I had your writing ability I look forward to more updates and will be returning.
Burn Xtreme
Chirp, Chirp. . .
Added both. It's a shame it takes a statistical SEO experiment for me to post a non profit Tweet. JK
Thanks for running test; look forward to the results.
Done. It'll be interesting to see on Open Site Explorer who passes the most link juice via a link.
Tweeted!
Tweeted!
I'm in with a tweet Mr. Dover.
Tweeted it !!
This sounds like a great little experiment! However I think the results are going to be slightly out of whack though, people are going to be more likely to tweet this than post it on their site. Still, I'm looking forward to the results...
Tweeted - Cannot wait for the results of this one.
Tweeted it.
I've also shared the link on Twitter.
How will you manage the fact that, perhaps, most of the people will share the link on Twitter but won't link to the website using the "Link" option?
Thanks, nice article. I'll tweet to help organizations address hunger SeeYourImpact
Does your twitter share option add to the counter down the bottom? Would be nice to be able to see how successful this campaign is with a tweet counter... Also, if there is a way to track incoming links at the same time that would be equally cool.
I don't have a twitter button set up but I am tracking this with backtweets.com. With regard to the non-social links, I can't track the links in real time but I can after we crawl them with Open Site Explorer and by seeing which sources send traffic via Google Analytics. Its not perfect but its the best we got :-)
Topsy is your friend: https://topsy.com/seeyourimpact.org/explore/sector/hunger/ending-hunger-sierra-leone/
I have one question though... Why the other page, the linked one, doesn't appear in the top 100 results, although it has 290 backlinks shown by YSE. it looks more like a penalty or something, since it's outranked by at least 100 pages that have at max 10 backlinks...
Can't wait for results have tweeted!
Add a tweet via tweetdeck .. posted this info on other SEO group, so get more tweets and clear results
Should be an interesting experiment. Though I wonder what the actual impact of social media has on rankings. I hope it's not too great for that would be bad for small businesses who won't have as many followers and retweets as more national companies.
Tweeted - hope results come soon!
Great idea...the results should be interesting!! Not exactly scientific but cool ! ;)
Tweeted! :-)
I wanted to look at the long term impact but now both URLs return a 404 Not Found :-(
Dopne too, Danny...
and we too await the results....
:-)Jim
how many tweets i can do for my company per day ? should i put short url of website to get traffic in every tweet ?
I have been interested in this experiment for a long time and am happy that seomoz can get volume to test it.
BY THE WAY - the site is ranked #1 on BING
Tweeted, though I note that this article is now #11 on Google for "Hunger in Sierra Leaone" and the charity's site is nowhere to be seen.
I've just tweeted - will be interesting to see which one wins.
Has anyone done a full analysis on this?
Hi Joe,
I don't work at SEOmoz anymore so I won't be continuing work on this project. I believe Rand is finishing it up but you can see the raw results here: https://www.dannydover.com/share/tweets-vs-links.html
Rand just spoke at the be-wizard.com event this weekend and showed a slide that showed a link that had been retweeted over 500 times performed better than a similar link with many backlinks
Tweeted.
Shouldn't you make the bit.ly link clickable to even things out?
;-)
Tweet Sent. Looking forward to seeing the results!
Done via Tweet.., same here curious which will have biggest impact
Tweeted, - excited what the results will be..
Tweeted. Looking forward to seeing the results!
Have tweeted plus tweeted the actual article. Always good to run experiments. Like you say 'only hints' but sometimes all we get to work with is hints, so we take what we can get :-)
tweeted and blogged. Looking forward to the results.
Done! Looking forward to results. Let's go traditional link building! :D [removed link]
great idea, i'm in.
Tweet
Tweeted!