The second thing that had me thinking about the relationship between social media marketing and SEO was a line from this piece by Danny. "Here's the thing," Danny writes. "Social media marketing is not SEO."
It certainly seems that search engines don't pay much attention to some of the traits of social media right now (such as votes on comments or even votes on posts.) Social media marketing can take place without much initial thought to search engines. When you have a piece of linkbait weaving its way through the pub-crawl of the social media giants, you don't usually pay much attention to search. Similarly, an effort to improve brand recognition or build a social media community isn't inherently tied to ten (or however many) blue links on a white background. Often, people are disappointed in the SEO results of an SMM campaign. While we've talked about creating relevant linkbait in order to improve real search engine rankings, the fact remains that a lot of linkbait efforts are a bit off-topic, unfocused on acquiring relevant anchor text or on acquiring links from related, authoritative sources.
It's hard to work out whether search engines currently understand or care about the voting systems and other features employed by most social news sites. If a piece of content receives a lot of social media votes, it often ranks first for the terms it uses, but is that ranking due to its social media success (in raw votes) or due to the number of people who have linked to the piece? Do search engines understand and care about the difference between two similarly-structured stories on the same topic, one of which was dugg 1000 times while the other received 4000 diggs? I doubt it: if the story with 1000 diggs was more heavily linked to than the story with more votes, I would expect the story with more links and less votes to rank better.
To me, this is the primary reason why SMM isn't the same as SEO. As Danny went on to point out in his article, there is a really strong tie between them and you're best advised to take part in one if you're interested in the other. However, if a search engine really doesn't care about social media attention aside from the raw links the attention can bring to a site, it is possible that a socially popular item could remain badly ranked. All it would take is a lack of links. I've had it happen to linkbait. It's irritating.
Digg makes it easy for a search engine to figure out what is popular, both on its newly submitted / newly popular stories and on stories that have received many thousands of votes. Suffice to say, Digg isn't using Flash or any other clever tricks to display the most minute details of a story's popularity, including its comment counts. Search engines could use this information to determine popularity, just as we know they use links.
Within a social media site (let's take Reddit as an example), the comments on a post are usually more important than the post's listing and description. The comments aren't normally more important than the source of the post (which is usually an external website but sometimes isn't); however, they provide the only useful content on Reddit's domain. Is it viable that as search engines come to better understand the way social media commentary and popularity works, that they will better associate the commentary on social media sites with the posts to which they link?
If they chose, search engines could already show indented results that don't begin at the same domain, but which relate to or reference the same piece of content. While this isn't the traditional purpose of indented results, it wouldn't be too out of touch with Universal Search's goal of incorporating alternative results and grouping results in different ways. Obviously, Google would come up with a far better design for such a feature; mine is simply to illustrate my point.
.jpg)
If they chose, search engines could already show indented results that don't begin at the same domain, but which relate to or reference the same piece of content. While this isn't the traditional purpose of indented results, it wouldn't be too out of touch with Universal Search's goal of incorporating alternative results and grouping results in different ways. Obviously, Google would come up with a far better design for such a feature; mine is simply to illustrate my point.
A better understanding of social media would help when a search engine goes to rank the alternative, indented results. Currently, Digg often wins the battle for SERP supremacy, although we've touched on how that appears to be changing. However, it would be fantastic if a search engine could determine where the most heated, informative, and interesting commentary was taking place. Digg comments aren't usually worth the pixels they take up on your monitor, but with a good understanding of relevant commentary, a search engine could give a smaller social media site a better ranking.
StumbleUpon would give search engines far more of a headache than the other services, since their data is usually far harder to come by, unless you're a paying customer. However, the site probably gives an intelligent engine enough information for it to process.
Of course, multiple metrics would go into teaching a search engine to play with social media and change its rankings accordingly. These factors could apply both when a search engine looks at a Reddit-like site, or at a blog like ours that includes social elements. The top seven that come to my mind are as follows. I'm sure there are dozens more.
StumbleUpon would give search engines far more of a headache than the other services, since their data is usually far harder to come by, unless you're a paying customer. However, the site probably gives an intelligent engine enough information for it to process.
Of course, multiple metrics would go into teaching a search engine to play with social media and change its rankings accordingly. These factors could apply both when a search engine looks at a Reddit-like site, or at a blog like ours that includes social elements. The top seven that come to my mind are as follows. I'm sure there are dozens more.
- Raw up-mod votes. Of course. However, some stories manage to drum up a lot of controversy or discussion without receiving very many votes. Recently, a Sphinn story gained four votes and over twenty comments that disagreed with the post. I'll save you the link.
- Raw down-mod votes (if that number is available). I don't believe that relating negative votes with bad search rankings would be a good idea. Sometimes, negative votes mean, "this piece of content isn't good." Sometimes, they mean, "I don't agree with this." A disappointing number of times, they mean, "I don't like the author / submitter / title / subject." At least two of these motives aren't good reasons to penalise an item in a search engine. In fact, negative votes could be a good thing in terms of detecting levels of interest.
Buried comments are also all plainly available to search engines, both at Digg, Reddit, and Sphinn. While Sphinn makes it a lot easier to see how many down-mods a comment has, it's harder to find a number in Reddit's source code. However, both use style="display: none;" to hide buried comments, which is a pretty good indication of their unpopularity. - Number of comments.
- Number of different people making the comments. A thread where two people go back and forth for fifty comments is probably not as valuable as when multiple people take part in a discussion.
- Number of positive and negative ratings on comments. Some sites, including ours, show up and down votes on comments. Generally, people vote on comments far less than they do on posts. Five or six thumbs up on a comment is a lot; five or six thumbs up on a post isn't very many. A relatively high number of votes on comments within a thread would indicate a lively discussion.
- Links within comments. If linking out from a post is one sign of validity, then a thread where people cite outside sources could be considered a positive feature as well. To combat comment link spam, an engine could discount comments that include both a link and a high number of downward votes.
- How often content appears to be added after initial spidering. As Tom Critchlow pointed out in justFred's post, a search engine should be duly impressed if it returns to a page many times, each time noticing more content in a thread. This not only indicates participation, but an ongoing participation. That is, a community's interest has been sparked enough to keep on commenting for a long period of time.
While I may not agree with the premise, excellent post since it got me to register. :)
There are so many concerns when it comes to placing any weight on social data that I don't think I could ever see it being involved in search results and having it be beneficial to the user experience. Some of the issues would include -
I could go on and on, but I won't bore people in my first comment here.
I do see some benefit in using SM data in things like Google's Custom Search Engine - where the search results affect only a single user or a small population using search results from a community they visit/identify with. I rarely search outside of my CSE now and it woudl be great if my CSE automatically learned that content from certain SM users (maybe linked through DataPortability or OpenID) have been identified as "authoritative" sources and their content rises to the top of my customized SERPs.
"I could go on and on, but I won't bore people in my first comment here."
If you have more to say it will please the mozzers to read it, as you have some excellent things to say. Welcome to SEOmoz!
"Very rarely are the most "informative" comments rated highest on SM sites - instead, the funny, witty Seinfeld like observation seems to garner the most upticks"
That is a great point. Categorization (informative, comic, educational, product review, etc) on the type of article with more votes could add much more value to search results than just amount of votes/thumps/comments/people commenting.
And that is why 'niche' actor should be added to the factor list btw, do you agree, Jane?
Britney Spears is more likely to hit the front page than SEOMoz beginner's guide (let's take it) :)
Obviously. I had no reason to mention that because that's inherent in search. The Beginner's Guide is not competing with Britney Spears for the same search terms and a search engine would would be highly unlikely to compare a Britney story and an SEO guide to begin with. Unless you've searched for "Britney Spears SEO", in which case your job is somewhat like mine was in late 2006.
LOL... OK, my example was kinda extreme. What if we compare "SEO tips for beginners" and "10 examples showing that all SEOs are bloody spammers". I guess the second post will get more successful in SM.
>>> What I meant to say is that with some niches SM reaction should probably be taken less into account than with other niches.
Yep, it will be more successful in social media circles, apart from maybe Sphinn and PlugIM, which is where a smart search engine could give more weight to social media outlets where it's obvious the community at those outlets knows about a given keyword, keyphrase or topic, just like search engines do now in determining that SEOmoz knows more about SEO than Perez Hilton and Perez Hilton knows more about Britney Spears than does SEOmoz. Sphinn members know more about SEO than do Diggers.
I don't see how "SEO tips for beginners" and "10 examples showing that all SEOs are bloody spammers" would compete for rankings. If they were both attempting to rank for "SEO" alone, that would be possible, but both of those posts would surely be receiving traffic from longer-tail queries? Maybe the second more than the first.
Again, I'm not suggesting that entire algorithms be overhauled for a social media metric. I'm suggesting that, along with every other factor search engines use to determine what searchers would like to see, social media data could be used.
Jane, we have already got it: not completely based on social media. You seem to be repeating that again and again. I already feel if I add something to the discussion, you will think I mean that you suggest that Google ranked sites only based on social media (which you do not naturally). :)
And yes, I agree that a smart search engine could be able to tell that Sphinn members know more about SEO and despite the fact they are in minority would take their opinion into account. That is still algorithmical ranking based on the niche.
I also agree that they are doing something like that already but you did mention other things they are doing now, so this little thing might also be included.
Again, I never said that these things should determine search rankings. I said that they could be taken into account. Just like links don't really determine the best content, but are taken into account when determining search rankings.
Jane, your article is great. Brings up a lot of good points to the discussion.
This is a great post by Jane & one I wish I'd commented on sooner, especially as hagrin has touched on (more eloquently than I would have done) one of the things that I find most fascinating about this: using social data to drive personalised onr vertical/thematic search.
Danny calls it Search 4.0, we call it 4d Search - whatever, it's intriguing. I also think that it is one of the most interesting aspects about the whole Yahoo!/Microsoft thing.
Between them they have hundreds of millions of email addresses (I know there must be duplicates, but it's still a whole load more than Google has with Gmail); Yahoo! also has some nice social sites (Flickr, delicious etc..), where people constantly bookmark stuff they find interesting, that they've so far failed to monetise.
Tie all of this together and you not only have an opportunity to provide truly personalised search, but also, and I think this is a big factor, truly targeted display. And that may well be worth $45 billion.
I don't see much value for the search engines in receiving most social data.
There is already a "social" system Google uses for news.
It looks at how many news stories exist about a subject and applies time to the equation.
If 1000 newspapers are reporting a story at 8 am, it will likely be seen on Google news at 8 am (and in the news one box results).
It does pretty well too, that system.
The quality indicators availible to Google are way better than what would be availible from a site like Digg, even if one story had 11000 votes and another only had 2 (or was not included at all on Digg).
Socal media sites are not universal, therefore their usefulness to search engines could not be universal.
Google can universally apply it's current system of counting stories and comparing the time line of these stories release.
It can not universally apply social site data.
Totally agree on the not universal bit.
But if they made a social media search (like images, scholar, etc) then they could test test a few combinations of those metrics.
If it doesnt already exist, you could make the first SM search widgit. Just take your "official sources google gadgets", and put in the URLs for all the social media sites.
I don't know who would like it best, SEM's, average joe's, but it would be super easy to modify.
Evan Jerkunica
"Five or six thumbs up on a comment is a lot"
Amatuer. ;-)
Nice post Jane, I think it's a very intersting theoretical discussion but I'm not sure that the search engines (or at least Google) will start actually counting the number of thumbs up/down content gets across the various platforms. After all, how many diggs is 30 thumbs up worth?
I think the element you mentioned that they are most likely to use is the sentiment analysis coupled with the number of comments.
Why rely on people to tell Google whether a comment was positive - why not let Google use sentiment analysis to determine that for itself? Once that's in place it could easily use IT'S OWN thumb up/down weighting per comment to help it decide whether the public opinion on a post was positive or negative.
Sentiment analysis is the future imho - once the search engines can crack that effectively then everything changes.
"After all, how many diggs is 30 thumbs up worth?"
Nice point, my sentiments exactly.
I like it when a post encourages me to think... Two serious concerns:
1/ Social media marketing is a popularity contest. There are plenty of examples when a truly good piece is left without attention. To get attention there you should either be promoted by Muhammad Saleem or be a problogger (who I guess doesn't suffer from Google traffic lack)...
2/ Social media is already heavily manupulated and spammed (this resulting in SEO hate that I of course do not support but then again there is no smoke without fire...) If G or respected SEOs hint that social media success influences organic search results, no one will be able to stop webmasters from paid-social-media-campaign (like now no one can stop paid linking campaign).
That's exactly what I was thinking, Ann. If thumbs up start to matter, we'll see legions of Indian shops offering to vote your comments up for $0.01 each.
Or enterprising webmasters will just adjust the thumbs up count in their MySql database. How many blogs have captchas? Or would care if people had multiple sock puppets just to thumbs up a comment?
The problem is that social media already does heavily influence the organic SERPs. It just has a short run time. Post a story to 30 sites and you'll be in the top 10 for a few days. But if the social sites don't keep buzzing about it, it starts to drop.
ok, now I am scared!
Disclaimer: dear citizens of India please don't get offended. This is just a joke. I for one never get offended when Russian webmasters are called "bloody spammers and hackers" :)
Thumbs down to you, Mr. Racist P.
Anyway, if it indeed becomes an opportunity to grow traffic, it is going to happen.
All Search engines or ranking methodologies start with the assumption that all people will do good. However, that changes over time and then, they do more of spam control......
Just the fact of life....Deal with it.
I think you're spot on there Ann - it would be a slippery slope to go down if search engines were to have a 'popularity index' in their algorithms. It would it be open to abuse as you explained.
Also something that's always bugged me about the 'social media revolution' is the amount of timewasting it attracts. Now I don't want to downplay how significant social media is in terms of the way we interact online, but often, whether it's facebook addicts or hardcore Diggers, I can't help but feel that a large part of the active demographic are people with nothing better to do with their time who get sucked into endless wasted hours online, regurgitating other people's content just to get some 'online street cred' (which by the way is inversely related to real street cred!). My rule of thumb is I like social media functions that help me do the things I already need to do, but in a more innovative and useful way.
So this skewed demographic is part of what opens to door to abuse - aside from spammers, it would allow those who spend their lives on social media the chance to bump up their friend's rankings while real authorities in their field may suffer simply because they haven't got time to go and canvass for votes. For this reason I don't see search engines giving this much importance, which links back to feedthebot's comment about 'universal usefulness'
Surely search engines already 'understand' enougha about social media using their criteria of useful, unique content and links? The more contributions a blog post, say has had, and the more 'useful' those comments have, the more Mr. Googlebot will see a post as an authoratative source...no?
Sorry for the verbal diarrhoea - I'm an SEO newbie and trying to get my head straight on all this, would very much appreciate people's opinions on this stuff...
I think this is one of my favorite posts you've done, Jane. I totally agree, BTW, and think that the Google "social graph" data suggests that relationships and intent of links and content and mentions is an area where Google (and all of the search engines) have a big opportunity to expand their abilities.
Next time you hang out with some engineers, you should totally bug them about this :)
Agreed.
As I've said before many times, the search engines really don't get Social Media — either as a driver of traffic or as a vehicle for dialogue.
Additionally, this article has helped pulled together some ideas of mine that I've been mulling over for some time.
And given that Jane was the catylist, I think I'll be sharing my ideas here on SEO Moz shortly...
You have to wonder if they aren't already in the process of using social media as a quality factor.
Very interesting stuff.
I have a hard time imagining that the work spent by Google trying to divine the importance of a post on a social media site based on popularity (thumbs up/down for example) would be more accurate than the numer of quality inlinks to the post. Juicy content is always likely to be linked to heavily and thus it's easy for them to assign importance using external factors.
They would also have to engineer for each social media format which seems sort of un-Googly. I don't imagine that they will pick up on social media signals unless there are specific results that could be improved by the adoption of new social media specific signals.
Great post Jane!
The term SMO, Social Media Optimization has already been coined, and by the looks of it is set to become a standalone component of web marketing, slotting in next to SEO.
The algorithms strive to emulate human response, and by studying trends and response patterns to given "bait" fed ino social media sites will evolve and learn. The advances in AI, coupled to this kind of data could lead to fairly accurate weighing of importance of inputs from the various social media sites and services.
We ain't seen nothing yet!
hiiiii
i hope you doing well.
actually i want to take 1 suggestion that, i have one website name www.thesariskapalace.in !!
i want to change this website layout. So please suggest me if we change this website layout then what search engine ranking affected.
please send me your positive feedback.
Thanks & Regards,
seemant
Re-reading the post, it seems like I'm suggesting that this would be an awesome idea that the search engines should implement and one that should make a bit impact on results. That wasn't what I was meaning at all... although I did try to get that across in the last paragraph by saying,
The popularity contest of links could be supplemented with the popularity contest of social media, especially if SEs found a way to harvest SU data, as I feel like that's more of a balanced community. Not entirely balanced, but more appealing to "regular users" as opposed to Internet enthusiasts and geeks :)
When I ventured to write that post, I started with the mindset not of making popular media more popular, but rather making the spam invisible. By spam I don't necessarily mean "buy blue widgets at bluewidgets.com", but also "haha that's funny lalala"
Seperating the noise from the chicken (the content)
Am I the only one who likes to read the first comment or two at digg on a popular story? Some of those comments are pure comedy!
Sometimes (often?) I read the comments before I read the post. I do this at SEOmoz as well.
It's because I'm from the Southern Hemisphere. We do things backwards down there.
It seems that the biggest obstacle here is that the system is too easily gamed; that seems true to a certain extent, but isn't it easy to game the system of inbound links?
The question is whether or not the search engine engineers are smart enough to figure out a way to distinguish between organically popular social media and spammed social media.
I don't see why they couldn't figure it out.
This is a really thought-provoking post, Jane. Thanks. Did you see the Vanessa Fox post on SELand about how Ask is launching a news search option that incorporates not only Digg submissions but also related, topical blog postings? It would be a huge win for Ask if they could be the leader in succesfully integrating social elements into search algorithms.
With such useful information, Lotus, you might want to include a link ;)
Thanks for the catch, Fred...I added a link.
The success on that will be given by search users, If people don't find useful to have blogs and other social media integrated on their search results, then they will switch to other search engine that will give them what they need. So I think the key is having the information on how searchers feel with the results they receive after a search. They will have to test, analyze and modify based on the results.
that 's a good post, as we all know that social media is growing up at a huge rate every day. which means that one day or another, they will have to come out with an algorithm in order to rank all those posts around.
I do agree with all the metrics and points that a search engine should look at, but how could it rank a discussion that speaks about the same topic on 2 different sites, saying that one site is more popular than the other one, but that there are more comments or thumbs up on the less popular one ?
Is it possible that Google will at some point have generated a separate index for social media? The way things are now they can’t include the number of “thumbs up” in determining the relevance of a topic. It seem that would turn every issue into a popularity contest.
If search engine rankings included the popularity of social media sites then some companies would just find a way to make money on that. I do think it is reasonable that Google’s engineers would find some way to incorporate some of the data from these social media sites. Maybe not over all votes for or against but the number of discussions. I hope I didn’t just write myself into a corner.
I am trying to say that I doubt the powers that be with the Engines would not be trying to figure out a way to become a powerhouse with social media. Its just a matter of time before a “digg” or “stumble apon” is joined by a “google preferred” or something like that. Then I could see the Engines have a different index ranking the social media sites.
Dear Jane ,
My self seemant from Delhi (India). Actually I m very Impress to you. Beacuse you have a lot off knowlodge about SEO. My dear Jane I have One Website Name www.thesariskapalace.in.
I am working last 4-5 months on this site, but i have not found more success. So i want to your help on this site. Pls suggest me what i od for this site high rank. I m also Working on SEO as a fresher. I have done B.Tech (I.T).
Mostly time I read your Blogs. realy you are ameging dear, I realy Impress you. If Possible then pls send me suggetion on my email id : [email protected].
www.thesariskapalace.in is a Hotel The Sariska Palace Official Website. This is a very Popular Hotel in India But when we type sariska palace in Google then this site comes on 21 number.
So Pls help me dear Jane...
I m waiting your Reply my dear jane...(My Email Id : [email protected])
Take care bye........
Welcome to SEOmoz Seemant
Your website uses a landing page, which is probably the main reason you are not seeing much success in Google.
When Google "looks" at this webpage it only sees this...
"The Sariska,The Sariska Palace,Hotel The Sariska Palace in Sariska,Jaipur,Rajasthan,India,Sariska National Park a text decoration none color a hover text decoration none color CA"
The above text is what Google sees and it will not see that text as very useful or relevant. The good news is that this is easily corrected.
I suggest you read the SEO Beginners Guide, and I also would recommend going to the Google webmaster help group, where you can list your website and ask for recomendations.
A great way to ask would be like this...
"Hello, I am from India, and I have a website for a hotel that is not doing very well in Google. Could you please give it a look, and let me know if I am doing anything wrong?"
To help you understand why your website is not ranking well for your terms you may want to read this or watch this video (it has specific ideas that can help you).
Note: I believe Jane understands that she is "dear"
The third or forth time you say "my dear Jane" may cause uncomfortable feelings.
Also of note, the same thing applies to leaving your email. Once is enough, and your website, again, once is enough.
Repeating things rarely works and will often be seen as a distasteful thing to do. I have made the same mistake myself. :)
Take care and good luck!
So who is this "Jane" and where can I find her profile page?
Feedthebot - Agree with everything you just said. I'd just like to point out that culturally, the language that Seemant used in addressing her would be completely appropriate in India. He was just doing the needful.
Cultural differences are amazing, aren't they? I've always found it really odd to be called "dear" by anyone but my grandmother.
I don't really need to add anything to that. Thanks Pat!
feedthebot, I really appreciate your sweet and humble reply to Seemant's comment. A big thumbs up for u ! :)
i try to send feedback