Today Google announced the release of a new social feature: +1
Read more about the launch from these in-depth blog posts:
- Danny's ever-excellent coverage: https://searchengineland.com/meet-1-googles-answer-to-the-facebook-like-button-70569
- Official Google post: https://googleblog.blogspot.com/2011/03/1s-right-recommendations-right-when-you.html
- Official Google page: https://www.google.com/+1/button/
Quick Summary
Rolling out across English Google over the next few days is a new "+1" feature that allows you to endorse URLs. If you're not yet seeing it in your search results enable it in Google experimental. Once enabled you see a little grey +1 next to all search results - including adwords listings:
Once you click a result you see something like this:
All of your +1 results appear on your Google Profile (e.g. Tom Critchlow):
I'm a really big fan of this from Google - they seem to be doing a lot of things right with social at the moment and this seems to be universally received as positive by the twittersphere. It's a lot of fun and ridiculously intuitive to +1 something and I can really see this catching on.
The Impact of +1 on SEO
So what's the impact of this for SEOs? Well I'm struck by the opening paragraph from the Google +1 page (emphasis mine):
The +1 button is shorthand for "this is pretty cool" or "you should check this out."
Click +1 to publicly give something your stamp of approval. Your +1's can help friends, contacts, and others on the web find the best stuff when they search.
Note how Google is emphasising right from the start that this is going to influence search results. Another quote from the official Google Blog (again, emphasis mine):
Say, for example, you’re planning a winter trip to Tahoe, Calif. When you do a search, you may now see a +1 from your slalom-skiing aunt next to the result for a lodge in the area. Or if you’re looking for a new pasta recipe, we’ll show you +1’s from your culinary genius college roommate. And even if none of your friends are baristas or caffeine addicts, we may still show you how many people across the web have +1’d your local coffee shop.
So the bottom line is that getting people to +1 your content is going to help you get more organic traffic from Google. Maybe even more/cheaper paid traffic too!
The Rise of Social SEO
Of course, for me this isn't so much a new direction as much as a continuation of the social circle work that Google has been doing recently. I'm a massive fan of results from your social circle - as I'm searching around these appear on a crazy high % of search results:
These social results pop up all the time and are immediately obvious and useful to me. The more that Google rolls out this integration the better imho.
Is this how Google are going to reduce the emphasis on links? Maybe.
Social Metrics Are Already Well Correlated With Rankings
I'm not going to go into too much detail here as we're still in the middle of gathering data and running analysis but here's a sneak peak from Rand's presentation that he's giving in SMX Munich next week. We've run a correlation analysis on a whole bunch of search results (~10,000) for a wide range of factors and there's some surprising results. Check out this graph:
It shows that Facebook shares are well correlated with rankings. In fact, comparing to other factors we see Facebook shares are similarly correlated to the number of linking root domains.
It's early days in the analysis and all we're showing here is correlation not causation but it's kind of surprising the correlation is so strong!
(Aside: I should point out a few things here - when we say Facebook shares we're talking about the aggregated number of Facebook interactions; comments, likes and shares as reported by the Facebook graph API. The full analysis will breakdown the different types of Facebook interactions in more detail. We should also say a big thankyou to Topsy as we have been using their totally awesome API to gather Twitter information)
In my opinion this is why inbound marketing is going to overtake SEO as the primary function of SEO professionals. Engaging across social channels to get links, shares, likes, comments and +1s is going to be the future for generating organic traffic to your site. Not just from Google but these channels are increasingly driving significant volumes of traffic in their own right.
+1 & Social Metrics Will Be Hard To Game
Previously the biggest objection I've heard from SEOs about user-generated signals is that they are easy to game. Well I'm not so sure. Think about how much information Google has on you and all the ways they can justify your profile is tied to a real human being account. For example - to show you're a real human being Google could look for the following signals:
- Gmail
- Google analytics
- Google calendar
- Adwords
- Google voice
- Google checkout
- Chrome sync
- Search history
- Google docs
- Google reader
- Youtube
- .... etc
Don't believe me? Why not head on over to your Google dashboard and see just how much information Google knows about you.
Still think it'll be easy to fake?
Combine this with some measure of author authority, which we know Google and Bing are looking at, and you have a pretty good picture of which accounts are influential and which are spammers.
Let's also not forget that Google are smart. I very much doubt that social signals will impact search results equally - some industries just don't have a strong social footprint. For these industries I think (hope) Google will normalise the impact and won't let the "fun" site outrank the "useful" site - they can easily tell which niches have a lot of social activity and those that don't. For the more mundane/commercial industries Google will fall back on the regular signals of links.
What's Next for Google +1?
Google are already talking about a new publisher button that you will be able to embed on your page to allow people to +1 content from your site - very similar to the Facebook like and tweet this buttons that already exist. Once you enable +1 you're also opted in to show this information on 3rd party sites in exactly the same was as Facebook buttons:
Here's a few other more speculative things to think about:
- Will Google create aggregated pages for the "hot" +1 content on the web?
- How will Google persuade regular people to create their Google profile page and add their friends?
- It seems like this is a very direct threat to the Facebook like button - how will Facebook react?
- How will +1 results impact Adwords listings?
- What kind of dashboard/analytics information will be available to publishers to see who is +1'ing their content?
For now, why not do us a favour and go give SEOmoz a nice juicy +1 :-D
So once again, all we have to do is...
Create great content
right?
Yes very true, it is similar to Facebook likes I mean really it is the same content if you make great content or have a great video then you will be able to utalize a large number of +1's.
I also can see they will have a great tracking beast behind these buttons, better then Facebook liek data tracking ;)
That's why I think +1 is a move in the right direction; this is voting for content in a pure form. It's like we are skipping the pretenses - in a system that's purely link based, links are votes, and only site owners who can link out get to vote. Now votes are votes, and everyone can participate in an easy way.
(Disclaimer: I'm not saying that links aren't still far more important than social signals. Yet.)
Since when is the popular vote the most accurate?
Great post Tom!
For us PPC specialists out there, there are a couple things to note: (and these are according to an email Adwords reps sent out)
1) A +1 in organic results gives a +1 to paid as well
2) +1s increase CTR
3) +1 does not affect quality score (and thus CPC) -- at least not yet.
However, since CTR does affect quality score, the +1 does have a roundabout effect.
Hey Roger! Thanks for outlining how this affects the paid side. I saw a lot of questions rolling around out there. I think it will be interesting to see advertisers (inevitably) add "Plus One this ad if you like us" into the paid search results. It's another social call for us to go after in the paid search advertisements that will affect our CTR, and like you noted, subsequently affect our QS.Interested stuff for sure!
No doubt you'll start seeing ads in the rotation with that exact call to action. Might even start happening before Google officially makes +1 part of the QS system.
I definitely see this happening on the paid search side of things, but I am not sure how good of an effect it wil have on actual clicks.
I would think a bunch of paid ads with the "Please +1 me if you like me" line is going to come off as spammy and will probably deter people from the actual act of clicking on the link.
I agree, it would look spammy. It is definitely something interesting to test out.
Its a big part of your word count too
Actually just got an email back from our Google rep and she said...and I quote:
"Ads that direct users to click on Google +1 buttons or ads that imply or directly promote offers, prizes, monetary equivalents, or services/deals in exchange for +1 clicks is not allowed."...of course we will have to wait and see.
@Joanna - The idea that, lets say the site had been +1'd 50 times because of various content on the site, if that indicator showed on the PPC adds we'd see a major uptake in PPC clicks. That may be the core motivator for Google, if their Google AdWords ads had a social element, had a "18 of your peers found this service useful" element, can you imagine the impact that would make on search?
I'll be curious to see if you +1 each URL independantly like on Facebook, or if you +1 each page of a site and it goes to one larger pool of +1's for that domain as a whole. If you get a bunch of +1's on your Blog articles, but not on your homepage, then your PPC ads would not get the credit from the +1's to say "16 people +1'd this advertiser" or whatever.
It's all really really cool all the same. Should make for a very interesting #SEOchat.
Great info - thanks for adding to the discussion Roger. Useful to know.
I spend about 1-2 hours per day on Adwords. Once the customer reaches the goal page, I am going to tell them to hit the +1 button. This will probably have no SEO benefit, but at least my competitors won't know what's hapening for a while.
I missed something in Google's explanation of the PPC site - is the +1 on the URL, or is there a separate +1 against an ad? It sounded like the URL gets +1'ed, and then both organic results and ads targeting that URL would display it, but I tested that and the ad was unaffected.
The +1 should transfer from organic to paid based on the URL of the landing page (and vice versa). I just tested it a few times and it worked.
I think the PPC version has a trailing-slash and the organic doesn't, or vise-versa, as I look closer. Guess I didn't canonicalize my paid URLs. Oops.
My first thoughts: another button we MUST add??? when does this stop?? Websites look more like christmas trees than back in the 80s!! Might as well start using BLINK again.
Anyway...
Hi Tom, I am not going to contradict or praise your conclusions.
I substantially agree with you but I need to ponder a little more all this +1 stuff.
Just one note: I believe that the +1 effect on rankings is going to be clearer when the button will be available for the sites and when Twitter shares will be included (and substitute practically the function of Buzz) Just then we will see a wider consequence of this social signal.
Also, if I remember well Google doesn't count Facebook Likes as ranking factor (more here: https://networkeffect.allthingsd.com/20110308/googles-approach-to-social-it-wants-to-find-the-one-true-you/ ), therefore I see it as the first real attempt by Google to counter attack Facebook on its own field. One thing is sure: this 2011 is really an exciting year in Search.
Google doesn't have access to the Facebook firehose, which is why they cannot count Facebook Likes. Bing, however, does take this social signal into consideration in indexing and possibly rankings.
At SMX West, Mike Cassidy of Google said that Google is using all the sources they can that are public and crawlable to influence ranking. Well, looks like Googl +1 is one of those influencers.
I agree about the +1 button. Seems to me that +1'ing a search snippet doesn't make a lot of sense. A +1 button on the content page, makes a whole lot more sense.
A very good point - we need to be careful about correlation and causation with Facebook. It may just be that sites that are popular on Facebook also tend to gain lots of links, which makes sense, so the data will be correlated - but Facebook interactions themselves aren't directly causing improvements in rankings.
Hard to game?
You mean easier to game for people who don't know SEO but can spam social media?
A very poor move by Google and a sign of their struggle to bring originality to the search battle.
A good day for SEOs, a dissappointment for those looking to see Google put up a fight against Bing.
Why put social voting at the search result level BEFORE someone even sees the website to know if it's worth "voting" for?
And how many people will hit the back button to go +1 a page?
More importantly...how long till bing introduces +2?
The end goal of this new +1 will to have buttons on all websites on the web, similar to Facebook like buttons but with quality tracking ;)
That was my first thought. By putting a +1 button before anyone has even seen the site, surely the only people who are going to use it initially will be the spammers and website owners that Google are trying to avoid?!
If i'm truly enjoying a website I will be digging into it further and bookmarking it, the last thing i'll be doing is hitting back and +1ing it (amazing how quick it is to invent a verb!)
I understand that soon it will be implemented in the same way as a Facebook "Like" but to launch it in this way is surely giving out the wrong message from the start.
While you may not necessarily want to click on the +1 prior to visiting the website, you will still be able to see others that have +1'ed the page/site. For this purpose, it will be useful. But until they enable it to be embedded in a similar fashion to the Facebook "Like" button on the actual site, it may not have a wide-spread adoption rate.
A very impressive post indeed. Given its public nature I think +1 is going to be abused left and right. Welcome the SEOCIAL era!
I agree; +1 will be abused, and Google will know pretty quickly which accounts NOT to trust or put value in when weighting future +1 votes. As an SEO and someone who cares about how my votes are counted, I'm going to be very picky with using the +1, and only endorse the things I'd really want to see in my search results and could back up as great resources for a given query.
Great post Tom!
I don't see myself using it to be honest. am not the pessimistic type infact far from it but this whole +1 thing reminds me of https://fugitivus.wordpress.com/2010/02/12/screw-you-google/
LoL same here. I also have something like that at https://chamberlain7th.wordpress.com
I think the math is going to be surprisingly easy. The same way they weigh things like link to root domain ratio, they're going to weigh +1's to other interactions ratio. So if you have clicked 10,000 +1's, but you don't ever comment on YouTube, you don't ever use gmail, you don't have any of those other interactions that Tom talked about above, then you're going to have numbers that indicate to Google the kind of user you are.
The average user might +1 something two or three times a week, if you create 20 Google accounts just to game the system, you'd have to go out and also do random stuff, and I mean REALLY random stuff to make your ratio's look "average".
The sheer volume of energy that would go into gaming that system would be so much better spent creating content worth +1'ing.
Now, what I do expect, is the rise of the Private Network, which was something shared during the Distilled #LinkLove conference. I have a private network on Facebook now, where I've invited other SEO's that I've met and that I think have a similar business ethic. We, in private, talk about SEO issues and we're free to ask questions we might be shy to ask about somewhere that'd have our name on it.
The private networks allow you to share your content with other SEO's or even just other bloggers etc. So what I suspect will happen is, legitamate people with real Google Accounts will band together to ask for +1's from their private networks. So whenever we create something, we'll also whore it out to those channels and say "hey can you +1 this?" much in the same way we might say "hey can you retweet this?" now.
It's all really cool. The gap between the Social Media consultant and the SEO consultant is closing every day. I can't wait to see what SEO firms look like 12 months from now as we each define what level of Social services we want to offer.
We're in uncharted waters, and I'm loving every minute of it.
Even if Google has measures to assess whether profiles/people are 'real' or not, +1 will still give a huge advantage to bigger companies with more employees who can (or are told to) +1 their search results. - Jenni
I didn't even think of that but it's a good point, I hope there's some way they can stop that from happening... always the little guy that takes the boot :(
I join you in congratulating Google on their commitment to integrating social signals into search rankings. However, make no mistake about it, intelligent people will do everything they can to come up with a scalable way to leverage this to their advantage. You cite the effort it would take to create the accounts and "go out and also do random stuff", but isn't that the kind of thing computers excel at?
It's easy to imagine someone making the a similar argument about links years ago..."well, to game this link based system we have to go create loads of websites and then publish content to all of them...the sheer volume of effort just isn't scalable!"
Alas, we all know how that turned out.
Again, I applaud Google's effort on the social front and think +1 is very interesting. I'm cautiously optimistic about the future of our craft.
Meh. Still. To post a +1 you have to have a Google account. Thus even just the age of the account can be taken into account when weighing those factors. If someone wants to create 10+ Google Accounts, go be active on YouTube, go send and recieve mail in Gmail, talk on all 10 accounts in Google Chat, etc, to APPEAR like they're the average Google Accounts holder... I mean... thats crazy.
Posting a link, writing a blog, that takes minutes, this takes months of "pretending" just to add one more +1? No. Not going to happen.
HOWEVER, we've already seen folks buying +1's via mechanical turk, but thats no different than buying retweets or facebook shares, thats nothing that doesn't already exist.
I have a legit google account, jet idont use chat, idont tweet, I dontuse it much at all, are you saying just because Ionly use it for adwords & adsense,if one day I decide to give someone +1 it will be counted as spam? Common, for many people like me twiter,facebook and google is not the center of the universe, I got life for instance.
Agreed, I just log into webmaster tools and analytics occasionally. I'm not really interested in anything else Google has to offer, and I'm not responsible for ads in our company. - Jenni
The problem with the 'fake Google account' argument is that it will probably be the least scalable and least effective way of gaming this +1 stuff.
Why bother creating numerous fake accounts (which would probably be easier than you think) when you can go on mechanical turk and pay people a nickel to +1 your content? Or, you could join a group who have a deal under the table to help promote other people's content +1'ing each other's content. Or, then again, you could just incentivize your visitors to +1 you on Google in the form of a promotion of some kind.
Absolutely! Abused + Misused. The big difference between Facebook/Twitter likes is there have already been algorithms developed that can measure the influence of users. With this.. all you need is what, a Gmail and a Google profile?
To get the most bang-for-my-buck I would way rather game this system rather than build up authority social media profiles. It would still be FAR easier.
I don't really care who likes what. It's not something I care about at all. What I do care about is discovery & then comment threads on about 1-2% of blog posts I read. Oh, by the way, Google requires you to tag that +1 box on your site with a nofollow & while you are at it could you please pick up a gallon of milk on your way home for them too?
You're assuming Google hasn't developed an algorithm to rank influence and trust of Google profiles.
Actually, in an interview on ThisWeekIn Startups featuring Lior Ron of Google Hotpot about 3 months ago, Lior basically said they crunch data to gauge trustworthiness and reputation of users who edit Google Maps. See here: https://www.youtube.com/watch?v=zEC42W-UucA&feature=player_detailpage#t=2055s
It's a huge challenge for Google to develope an algorithm to get social search right, but I got my money on them any day.
And I'd bet they are wayyyyy further along than we think we known.
You're grossly underestimating google.
As the blog states - there will many methods in place to prevent this. They could even have implemented a system where all profiles younger than 6 months don't count, accounts that don't get weekly activity don't count, etc.
Good luck trying to game a multibillion dollar company that's sole purpose is to prevent people from alterting it's search results to their advantage!
I have tremendous respect for Google, but we do have to temper such sentiments with the realization that it's been how many years and they still can't find many paid links unless other SEOs out them?
It's great that Rand isn't going to abuse this, but I bet I could get A LOT of people over at Fiverr to +1 a tonne of my links. I think google will mainly use the +1 from within people's social circle and this will be another push for people to use google buzz.
I wonder what they are doing with postrank and how they are using all this social data in their social ranking factors.
"Will Google create aggregated pages for the "hot" +1 content on the web?"
Did Google just give themselves the tech infrastructure to replace Digg and Reddit?
Great point Cyrus - I really think they should have launched with that feature, to help encourage more +1'ing.
That's the first thing I thought when I heard about this... it's all a little to similar to those social media to be entirely removed from them
Pretty weak branding considering that people who look for "+1" will be getting results for "1" (dropping the +) until such point when +1 becomes authoritative enough to be seen as relevant for just "1". A sign Google departments don't really talk to each other? Perhaps they just don't care as this was mean to go public via social channels anyway.
No reason to fake google profiles, people will be more then happy to particlpate in schemes where they get something in returns for +1ing different URLs. Webmasters have proven this already - then bloggers - then twitter users proved it again - now anyone with a gmail account can become paid endorsers. Guess what, people like money and plenty are willing to endorse other websites for money. +1 take much less effort then putting a link on your preciously built website.
But on that same line, if it is so easy to +1 a page, then the value of that action would be next to nothing. If someone was to buy a +1 from someone, it would probably cost more it time to set up and manage ensuring that the +1 paid for was done as contracted for, and I doubt that a single +1 is going to cause great SEO benefit on a site with a budget large enough to consider buying links/endorsements.
I don't think the ROI would make sense, which is a good thing for WhiteHatters.
Can't wait for the "Google's +1 Blacklist" to be first rolled out and massive penalties for folks abusing the button...
Do you really want to see when your Dad's just +1'd asianbabes.com?
This is so amazing that I'm scared it's an April fools joke :(
Great Post but I've a question: How visitors can click on the +1 button before visiting your site? I don't think that they'll do another query into Google's search engine only for checking the +1 button...
That is definitely one of my complaints about it; SEO and geeks will use it but others maybe not so much. However, once Google allows you to hook into it from your website and elsewhere, then people will use it more and having it in the search results makes sense then.
An interesting point is if OpenSiteExplorer, Majestic etc. are going to be able to read the number of +1's in. Currently, it doesn't show a total count, as it depends on your network.
Dude, like I said above here: they will update the +1 after the fact (that's my lucky guess) where you "+1" a blog post for example on the website itself and your +1 will show up in search results later, probably the next day.
You don't need to lucky guess, Danny Sullivan had the inside scoop and confirm this feature is coming 'in months, not weeks':
https://searchengineland.com/meet-1-googles-answer-to-the-facebook-like-button-70569
Personally I wouldn't be putting too much weight behind the correlation between social media engagement numbers and rankings. I think it's a 'chicken and egg' scenario; pages that rank well are likely to receive high numbers of likes / comments both because 1: they are highly visible in the search engines, and 2: there's a high degree of overlap between ranking factors and what makes a page highly shareable (reputability of the source, linkability, quality of the content...)
Definitely worth of further research though!
Cue "correlation doesn't equal causation" - all too true in this case, but even if we assume there is no first order effect, which based on several experiments I've seen/conducted would seem to be a risky assumption, we KNOW there are many beneficial second order effects of SMM on SEO.
Specifically on the implementation side of placing a +1 button on SERPs--why am I supposed to like a page I haven't visited?
Seems to me that google rolled out the wrong code first. The +1s in the SERPs make sense only if there are +1 buttons on sites, the main benefit to the SERPs buttons would then be reporting what had been "plus oned" using the onsite buttons. The ability to +1 a search result doesn't seem very intuitive to anyone except those interesting in gaming ranking signals. Silly Google ;-)
I was just thinking that. In order to +1 a page you would have to know what's on the other side. Who wants to remember or even go through the process of hitting their back button until they get back to the SERPs page to +1?
Seems like it would increase bounce-rate, which, if truly incorporated into the algo, would be somewhat self-defeating!
I think they are onto something that will work and will be a motivator for users to click on one search result over another. Not too keen on the privacy implcations - as with most Google products (and FB) privacy is getting harder and harder to find.
Rand, can you share more about your graph presenting Spearman correlation coefficients with rankings?
Very soon, yes :-) The full ranking factors should be out sometime in May, and will include all sorts of data about methodology, specifics, etc. For now, I can say that the process is extrememly similar to what we did last year - https://www.seomoz.org/blog/google-vs-bing-correlation-analysis-of-ranking-elements
Great news. I hope this becomes an annual event.
Good post, I am glad you mentioned the mundane industries. These are often forgotten about when posts around Google social are wrote. I agree that Google will need to fall back to traditional link type factors for those industries that are not so "Social". When I first read the Google blog post, I thought it would be a bad move for them to admit it would affect rankings as it would be a party over on Mechanical Turk. But after taking a step back and having a think about this, I agree it could help weed out spammy profiles that are setup just to go around +1'ing everything for 0.5cents.
Going to be very interesting to see how this unfolds. Also, is Rand going to present the same data at SMX London, I can't make Munich.
There's one thing to consider: Facebook is about 5 times bigger then Google when talking about registered users. Gmail has roughly 100 million users, and they are not necessarily logged in while searching – most other Google services have negligible numbers as compared to Gmail and certainly Facebook – which it's users are almost always logged in on or off site (vial the like button)
So the question is: can Google create a mass +1 voting numbers to affect SERP's? I am not really sure they can, I think that at the moment people are quite happy with the current like button and its massive integration into almost any website.
Honestly. It think this battle cannot be won by Google…
+1 is the like button of Google. It's just a matter of time when your FB login will be tied into you searching on Google and Search Results will vary depending on your social web.
Sorry if this is a bit long, but this put a lot on my mind.
First, this is too easy to game. I can already tell you that a company like Overstock or JCP is going to have every employee +1 their result. I can tell you there are warehouses full of spammers in India and Singapore who can't wait to offer the +1 service.
Authority profiles? Give me a break. 1,000 spammers in India can have real profiles they use everyday, and they can sell their collective +1 votes and make themselves appear organic by simply +1ing random and popular stuff. So easy to game, even a caveman could do it.
And like someone said earlier, I don't care in the slightest what other people like. Do you like Rebecca Black or Justin Bieber? How about Farmville? Well, LOTS of people like these things, so should Google show them to you more prominently in search? NO!
Think about how things get spread on the internet anyway. What percentage of people who like something ever click the like button? What percentage of people who like something ever share it with someone else? Most normal, average people simply don't do this very often.
In my opinion, the thing I get shared with me most often is funny stuff. What gets shared and liked by the masses is what they see as entertaining, not useful. So now Google will be like the news, a bunch of garbage no one really cares about. The Onion and Daily Show are going to dominate the SERPs. Get ready for Kim Kardashian's important tweets to grace the search results regularly. That's what happens when you open up ranking factors to the masses, you get what the masses like, which isn't all that useful or meaningful.
Google can, must and will find a way to have people influence rankings, I firmly believe that. The era of links is ending and we all know it. I think the +1 is a step towards a more social SEO, but the future of SEO is just getting people to like, tweet, +1 or share my content, then a lot of my websites, and a lot of yours, are utterly screwed.
As to the statistical test of the Spearman, I still take exception to the fact that anyone is calling something less than .3 a strong correlation. You can look this up in a quick Google search if you don't know stats. Nothing is considered strong until .5. Plus you not only get a correlation coefficient, but you should also have a 2-sided significance score, right? What is the significance of this test? I don't recall seeing that reported.
I have no problem with you running tests, but I think those of us who have been in statistics would like to see something perhaps a bit more academically credible. If I were submitting a study to a peer reviewed journal and I claimed a strong correlation based on a .3 or less correlation coefficient and I didn't mention the significance of the relationship, I would be laughed out of town.
May I make a suggestion for your testing? First, sample sizes of 30 are the minimum that most typically accept in studies like this. I know you look at 10,000 different serps and that is good, but why not set the serps to display 30 results instead of 10? This brings the sample size of each serp you examine to an acceptable level.
Next, separate your nominal and scale variables. Nominal are the yes or no ones, like, keyword in title? Keyword in alt tag? Keyword in H1 tag? Keyword in strong tag? Keyword in URL? Then run a T-Test between these and their ranking. Do any of these seem to impact the ranking? Are the rankings of those with keywords in the title tag significantly higher than those without a keyword in the title tag?
Then I would look at my scale variables, how many tweets, how many likes, shares, links, etc. You have already run a spearman or pearson it seems, so instead of correlation tests, let's run causation tests. To do this, you need to convert the scale variables into nominal, ordinal or interval variables. For example, say in all the URL's you examine the fewest tweets to a URL is 0 and the most is 100. Convert the variable into categories, 0, 1-10, 11-20, 21-30, etc. Then instead of getting the number of tweets, you give them a number based on the category they fall into.
Now you can run a One Way ANOVA. Look at the average rankings of the different groups. Is there a statistcially significant difference between the rankings of pages with 50 or more tweets vs. those with none? Maybe you could convert the data into two categories. Find the average number of tweets by median, say its 20. Make two groups, 20 tweets or more or less than 20 tweets. Then you can run a T-Test again.
I think statistical analysis like this would immensely build the credibility of these reports in my mind and the minds of any others who have done statistics in academia. Sorry for the long rant, I'm done now :)
This will be covered in more depth when the big data set comes out (by people more qualified than me to discuss it), but I'd like to address the point of what constitutes a significant correlation. In my experience, coming from the social sciences, there's no fixed number (like r=0.50) - it depends a lot on the number of factors involved. The more factors in play, in a multivariate model, the weaker any given factor is likely to correlate with the measure you're interested in, so the statistical bar for what constitutes an "interesting" factor drops.
Google claims to have over 200 ranking factors in play. It's incredibly unlikely that any single ranking factor correlates 0.50 or greater with ranking. That would represent a big imbalance in Google's models. So, if we set the bar at r=0.50, we're basically saying that nothing is interesting. Obviously, we know that the factors exist, so what qualifies as "interesting" becomes relative, mathematically speaking.
I know I'm stretching it a bit, but here's another example. Let's say you have a presidential election, and the winner needs to get a majority vote (>50%). In our (usually) 2-party system in the US, that works out fine. Let's say, though, that you have 200 candidates, many of whom are popular. With that 50% criterion, your country will never get a president. Mathematically speaking, I think that's pretty close to the situation we're in.
And that is precisely why I think significance should be reported and more than just the top ten should be looked at. Correlation is a pretty useless statistic in this case simply because of how many factors and variables exist. Correlation doesn't look at causation, it simply observes that lots of sites that rank well in Google are also liked on Facebook frequently. Well, do they rank because they are liked or are they liked because they rank? No one really knows.
I think if you look at the top 30 results instead of just the top ten and convert scale variables into cateogrical variables you could say something like "The average ranking of a url in the top 30 with 50+ likes was 2.6. The average ranking of a url in the top 30 with less than 50 likes was 20.6. By running a T-Test of these means we found a P value of .01, which means the differences in these averages is statistically significant and 99% likely not due to chance. This suggests that getting 50+ likes on Facebook significantly contributes to your ranking in Google."
Don't you think something like that would be more useful than a Spearman test of correlation? I think SEOMoz has the data to make this happen and should try something like this.
My understanding is that (as in the past) those stats will be explained in detail in the full analysis. Tom was just pulling some teaser numbers. I can assure you that SEOmoz has put a lot of investment into trying to find the best math for the job, but some of this is uncharted territory.
I gathered the data and ran the analysis above that Rand will present in Munich. All of these results are still preliminary and we will document the methodology in the final report, but I wanted to quickly respond to Tom’s points about the statistics. I agree that reporting significance/confidence bounds is a good idea where possible and we plan to include this information in the final version.
+1 on Dr. Pete’s excellent explanation on the significance of the magnitude of the correlations.
We used 30 results for each SERP for exactly the reasons Tom mentions.
Our data isn’t normally distributed, or even close to normal (rank is uniform and many of the features are very non-normal) so t-test and ANOVA aren’t appropriate (in some similar analysis on Google Places, I tried ANOVA and found the normal assumption wasn’t justified: https://www.seomoz.org/blog/google-places-seo-lessons-learned-from-rank-correlation-data). There is a non-parametric version of t-test (Mann-Whitney U) that I plan to look at when given a chance.
That actually makes a lot of sense. You wouldn't measure ranking as a scale variable because it is really more like an ordinal variable. At the end of the data collection you will have just as many 1's as 2's and just as many 2's as 3's and so on. Definitely not a nice bell shaped distribution.
I have two thoughts. Perhaps you are going about it all wrong. Instead of making ranking the dependent variable, why don't you make the independent variable? Then you make scale variables like tweets and likes be the dependent variables. Then you could run an anova perfectly well. What is the mean number of tweets and likes for all the url's in position 1 vs. all the url's in position 10. Compare the means along with the means of all the other ranks and see what you get. I think this can work very well because the number of tweets should certainly be more bell shaped in its distribution.
Another thought would be to not measure the rank as a scale variable like you are kind of doing with correlation data. Correlation only works well with two true scale variables. Rank is an ordinal variable. Why not convert your tweets and likes and everything you measure into ordinal or nominal variables as well and try a chi-square? You have a table with 30 rows, one for each ranking, then say 10 columns, one for 0-5 tweets, one for 6-10 tweets, etc. If you did 10,000 queries and got 30 results for each, that's 300,000 data points which is more than enough to fill that table. Run the chi square to see if the distribution is significant.
I would love to see data from these two tests as I think it would be more impactful than correlations.
In general, most of our variables are not normally distributed (tweets, number of shares, number of links, etc all have long tails at high values) so parametric approaches like ANOVA won't work. In some cases there are non-parametric alternatives that would be worthwhile to look at and I plan to do so (remember, these are just the very first preliminary results). Our data set is also structured such that pooling all the data together for analysis as in your Chi-square suggestion isn't always appropriate. The results in a SERP with lots of traffic and would have a higher number of tweets, links, etc then the results in a tail search with only a few searches a month. In addition, significance tests only have a 0-1 response, significant or not significant. We also want to know which factors are the most important, one of the benefits to looking at correlations.
Yeah I believe it is the Kruskall wallace (sp?) that is the non parametric version of the anova, that would be worthwhile to look into.
You know what would be really interesting would be to combine variables. For example, suppose you have keyword in title tag, keyword in alt tag, keyword in strong tag and keyword in url. You get a 1 if you have it and a 0 if you don't. Suppose a url has all of them. You could make a new variable that's a 1 if they have all 4 and a 0 if they don't. Then see how well that variable correlates to rankings. You could run a mann-whitney non parametric test to compare the means of the 1's vs the 0's.
That could work and be interesting.
"Authority profiles? Give me a break. 1,000 spammers in India can have real profiles they use everyday, and they can sell their collective +1 votes and make themselves appear organic by simply +1ing random and popular stuff. So easy to game, even a caveman could do it."
That's the big challenge Google has in front of them. I think Google announcing +1 is them basically saying "Challenge Accepted!" +1s won't be a major ranking factor until they're confident in their ability to rank the trustworthiness and credibility of a user (just like they've always done with links) and they're confident that it produces better search results for searchers.
I think Google is way further along than we think on this. See my post above linking to an interview by Google Hotpot lead Lior where he talks about crunching data to gage the trustworthiness of a user's edits in Google Maps.
Here is the issue. In order to determine if a listing is "good", one needs to click on the listing, visit the website. Assuming the website is "good", one must go back to Google, type in the same keywords, hope the listings are the same, then click on the +1 link on that listing.
Sounds very cumbersome, and time consumming for the average user.
Only the SEO savvy will use this function, and they know how to GAME the system.
Another failed attempt at the social internet.
I bet Google understands this and will also offer +1 widgets for the web sites
Dude, how many buttons are we going to have on our sites? Pretty soon we won't have any content, just like buttons...
I find that for most sites, people only actually use the buttons for a couple of social networks. If you have any buttons that people don't use, get rid of them. It can help speed your site up too. - Jenni
Hello guys,
I guess it is interesting to know how many users have a Google account. Do you know that (per country)?
Unfortunately I cannot find any information to that in the internet.
I think, the people who have one use it mostly for business or use Gmail. All of my friends (they don't work in the internet industry) don't have one. I cannot imagine if these users use the +1...
So, for me it is very interesting if +1 is a really assimilable with the Facebook Like-Button. I don't think so, especially in Germany. I think Social Media is not one of Google's best thing. When I look back on Google Wave or the integration of Twitter... hm, is it worth?
Best from Germany!
Kathrin
Please, wake up all SEOs who are spending all time behind link building.
Decor your home & than invite friends for dinner party!!
Initially, SEO need to focus on strong web page with unique attributes. Right??
I am really happy with +1 facility which will help me to refine & see accurate result during long trail keyword search.
How can I know that, how many +1 I receive on my own website?
I give this post a +1 .I can already imagine the impact of this on local. This is truly an exciting time to be involved in web marketing . The early adapters will truly benefit from all of these rollouts and shifts.
great article, thanks
Let the social games begin.
Hi noyellingds! I like your reply! But i think it became Social War! Well! It sure game for Social Media Optimizers.., Piece of cake it's going to be with Google plus...
I'm excited to see the future of +1s. I'm also curious to know how it'll work along side with Google Buzz - will it be another "ladder" to climb or just a replacement of the Buzzing of posts. It's interesting to see Facebook move away from their 2-share system ("Shares" vs "Likes") and now to see Google move towards a "2-share system" (Buzz and +1s)
2011 is already looking like a fun year for the industry!
IMHO the real desire of Google is to dismiss Buzz and use Twitter.
What makes me think so is that:
What does this mean for small businesses, the little guy? If this has a heavy influence on rankings, the little guy will lose every time. A large corporation with millions in resources and hundreds of thousands of customers are going to have a considerable advantage and be able to influence rankings very easily. Just because a business is smaller it doesn't mean they are any less relevant or offering an inferior product, but that is exactly what google is saying, if the +1 has even a small amount of influence on rankings.
These large corporations are also most likely the biggest spenders on Google's AdSense. They will continue to spend and the little guy will be forced out of the top rankings and either go out of business or be forced to spend more money with Google on AdSense.
Remember the Digg update that basically killed the little guy(and also killed Digg's user base)? In my eyes, this is exactly the same thing that is happening with this Google update, except the majority of the user base won't know any better.
For local brick and mortar business, I think this will be a huge + for companies to win online just by providing great customer service. Which is something they should be striving to do anyway. Now they don't have to worry so much about building local links, submitting to directories, developing citiations and all that other BS.
But ya, for small businesses that compete nationwide vs big brands it remains to be seen how Google will make the playing field level so that the cream truely rises to the top.
Nice article, thanks
One link having many +1 votes!!!!!! It could be one of cool tricks of Google to track spam.
I think +1 is another step toward personalized search results. I agree that the +1 vote will most likely be abused in order to drive pages to the top of SERPs. Social integration is becoming more and more important for SEO.
And the rich get richer...
This is such a bad idea.
Maybe someday there will be a social tool which can measure value rather than popularity. But I am not holding my breath.
It seems like this is a very direct threat to the Facebook like button - how will Facebook react?
TBH, I don't see it as a threat to Facebook.
With Panda Update and the +1 Button in additional to the other social signals, Google is doing their best to force people to create unique/genuine content AKA websites and to decrease how Links affect SERP and I think they are doing a great job so far.
I really can't see Google placing much value in this from an algorithm perspective at all. After all, the most +1's will go to the sites already at the top of the SERPs, so what would be the point? People still aren't going to dig down to the 5th result page and decide to +1 something (unless it is their site, I guess, and again, no value there). So it would be silly to give much / any cred to these values.
Where it WILL have value, however, and what it seems like Google is shooting for, is that you will see when one of your "friends" has liked a certain site, and therefore will improve your CTR because it is a trusted vote from someone you know. Seems like Google is going to be investing heavily in the Facebook / friending model, and that is where this will play a factor.
As everyone says, this is a HUGE change. Google continue to make the search more and more social.
I have a question:
- Does anybody knows how to have a button into a non-Google website?For example: we have in our blog a Facebook LIKE and a Retweet. I have to have this +1 button.
Do you know if it's possible now? I've read all the articles, but I couldn´t find it.
Thanks!
Not possible... well, someone did it (look at Search Engine Land posts of today), but Google has already banned it.
It will be available in some... months; or so Google said in its official note of yesterday.
Ciao Gianluca! Grazie per la risposta!
I want to have it in our blog! haha
Thanks,Fran
Yeah, about 10 minutes after I tested a +1 button in dev environment and was about to begin working on the error control for when the service wasn't available, the service was completely blocked LOL. At least it saved me the trouble of writing that code... for now at least ;-)
Google +1 nice working and also showing in Google Profile +1 option .
So Google would only take data from Google dashboard ey? Like Gmail and stuff? Then what about old people who still use hotmail, search with yahoo or bing? They are considered not participate in the +1 social seo then?
People with active Google activities and follow in with Gmail and stuff are likely to have website of their own and therefore they would +1 website that aren't in competition with them. While those people who use yahoo, hotmail, bing who normally don't have website/blog of their own would give a more objective +1, but too bad they would be out of the equation.
It's going to be an exciting next couple of months as we begin to explore these uncharted territories where the boundaries of search and social meet. Fasten your seatbelt, hold on tight, and enjoy the ride.
Totally agree that people will try to leverage this and then Google will end up changing how they use it - in Google fashion. I also agree that most people won't even notice that it's even there besides of course the geeks and SEOs.
Great post Tom!
So just like on Facebook where everyone went mad with FBML pages and banners asking their visitors to click "Like", or be in a competition if they "Like" this page... are we all going to end up having to spend time convincing all of our page/site visitors to +1 us now? If so, surely that defeats the object, we'll be spending time trying to come up with elaborate ways to get people to +1 us when we could have been doing what G claims it really wants, creating decent content!
Because that's all that will happen, the same good and bad content will come out at the same rate, it's just that all of it will include tactics and ploys to get +1'd.
I don't know... maybe I am old and I am been burnt by Stars, SideWikis, Buzzes and Waves, but when it comes to Google and Social I tend to be skeptical.
I do not deny that Social Signs are and will be an important element in the ranking factors, we are seeing so many signs proving this. But I don't know if the +1 is going to be a mainstream success:
On a SEO perspective I don't see in the short term a great SEO value in the +1. Nope, until Google won't provide APIs and Widgets for sites and applications (or maybe a complementary integration between Twitter and +1), this new social invention of Google will be just a very minor ranking signal, as it is quite easy to spam and because their numbers are not going to be statistically relevant. Then, well, it will be probably a totally different tunes.
But the most important question is: is it Google going to have a critical mass of Active Google Profiles, therefore of people seeing and desiring to Like... ops, to +1?
google +1 is just google answer to their slipping page rank. Page Rank has always been a sliding scale and facebook has destroyed google in this area by using share and like buttons. Every fortune 500 or .edu has a share or like button. Now that is some serious LINKJUICE! These sites will avoid linking to google unless they absolutely have to! Google saw the PR decline long before we ever got a chance to see it. I am surprised there isn't more debate about this. Am I the only one to have seen this angle?
I have a very basic question. Why would I +1 a website without clicking on the result. Even if I like it after clicking on the result, I would never go back to results to +1 it. +1 should be on the website itself not on the search listing (in Google toolbar; like side wiki or something).
Cool! Yet another* way to "like" a search result!
*Of course, I refer to the old-fashioned method, previously referred to as a "click". Now, we have both a "click" on the link itself AND we can show our friends we "clicked" (or not) by ALSO (or instead) giving the link a "+1"!
Glorious days to be a link in a search result ...
If a big corporation decided to "ask" its employees to hit +1 on their website and these were all legitimate google accounts that had the right information filled out and we'ren't necessarily connected to said corporation, then what.
...feels a lot like the alexa toolbar to me, only with EVEN more influence.
I can almost see the spammers licking their chops over this one. But I think you're right, Google must have some kind of measure in place to assure that websites are being unfairly promoted or demoted by +1. It looks more and more like Google is preparing to enter the social networking realm and they are getting their users ready for it.
Good Post! I am not in US and may be this is the reason why I cannot see the +1 button on my side. This seems to be very good information but… I had a discussion few days back about does Google take Facebook like as a ranking signal?
Let’s suppose Google do take Facebook likes as a ranking signal then if +1 feature will get success (seems it will) so will Google stop tanking Facebook ‘likes’ as a ranking signal? As Google didn’t have much good relations with Facebook!
Interesting, so I think big sites with huge social media presence will be able to totally dominate this new change.
For example you have a authority website with a few million Facebook fans, you can just advise users to +1 all your website sub pages in various niches, then they will rank for a various amount of keywords.
Imo Google should have tested this change in a smaller market first such as .com.au (Australia) before letting it go wild.
I agree with your concept but not in the execution of what you're saying.
Remember, these +1 results only effect people IN YOUR SOCIAL CIRCLE. So yes, that 1 Million +1's is going to be great, BUT if you're not in my social circle. If' I've never interacted with you via Google, or Gmail, or whatever, then your +1's are not going to impact my Search Results.
Now if they make +1's an indicator, and give them weight like a tweet or like a link, then yea, the larger communities who can get things mass +1'd are going to have some seriously large e-wangs to throw around.
This sounds to me like it will be affecting more than just your social circle:
"And even if none of your friends are baristas or caffeine addicts, we may still show you how many people across the web have +1’d your local coffee shop."
On the SERPs it looks like, for now, it's only showing you +1's from your social circle.
On folks sites, like a Facebook "Liked 156 times" button, that would say "150 other people +1'd this".
Yeah I know it is in your social circle but what I am saying is that any one with a large online presence will be even more powerful now =) I agree with the fact that if you promote this too quick then it will burn off similar to building links to fast/in bulk you need be smart/natural about it.
This goes along with what magicrob says above, and I love it. "All we have to do is create great content"
This is just one more reason to BECOME that purple cow in your industy and be worth talking about, worth following.
Influence is the new advertising. Deserve it, and you'll go far.
Hey Matthew,
Good to bump into you again.
I found this Q & A on Google that has literally disappeared in the last 2 hours, but I found this answer below quite interesting. It clearly says +1 is a ranking factor. Whether that is just for people in your social circle, it doesn't clearly state, but my thoughts are that it will (eventually) also influence non-personalized search results.
Does +1 affect my site’s performance in search?Content recommended by friends and acquaintances is often more relevant than content from strangers. For example, a movie review from an expert is useful, but a movie review from a friend who shares your tastes can be even better. Because of this, +1’s from friends and contacts can be a useful signal to Google when determining the relevance of your page to a user’s query. This is just one of many signals Google may use to determine a page’s relevance and ranking, and we’re constantly tweaking and improving our algorithm to improve overall search quality. For +1's, as with any new ranking signal, we'll be starting carefully and learning how those signals affect search quality.
Kevin
Hey Kevin! I imagine it will be a ranking factor, but I would hope that Google is smart enough to not overweight their new "innovation".
I worry about Social Signals from Twitter and such things because Twitter makes up such a small chunk of the web. We "social media elite" are, through our early adopter behavior, dictating to the search engines what is worth ranking #1. I'm not sure we should have that power.
Early adopters tend to prefer different things than the average consumer. So Google, in trying to give the best search experience possible, is slanting their view of "best" towards people like you and me who really get off on this stuff. While I don't mind being given that authority, I don't think we necessarily deserve it.
As long as house wives and Joe the Plumber, who both think Twitter is "silly", outnumber us the early adopting elite, I'd prefer an algorythm that helped them find what they're looking for, instead of finding the hot new Social Media trend that we're talking about.
When you weight the influence of one group of people, and that group has strong opinions and trends towards newer, flashy, innovative (or seemingly innovative) stuff. That weights the Search Engine results based on our very biased sharings on the web.
Joe the Plumber will never use Geckoboard or Freshbooks. Yet we can't seem to shut up about either, and that awards both with "web authority" over other, less Web 2.0 solutions.
Matthew - Excellent point. If search skews everything toward the early adopters your average Joe is going to find search to be useless and annoying. There is nothing more frustrating than searching for something and not finding what you want.
My question lies within whether Google can make this a huge hit or a slow belly flop. Sure we all like to click "like" but will we want to click "like" and then head over to Google to click "+1"? And without the button installed on websites, it seems a little useless at the moment. If I search for something, I'm certainly not going to give a "+1" until I view it. By the time I'm done viewing it and wanting to rave about it, am I likely to go back to the search page and do it? Adding that step backward is not going to make it very efficient.
Matthew,
You make a valid point that should be extremely important to Google when making these kind of decisions. However, we could make the same argument about the demographic that links to content (and the massive manipulation of that as well). By no means do I think +1 will be a non-personalized search ranking factor this year, but I am interested to see how it plays out.
My initial thought is, that if it succeeds, it will be an important indicator. Possibly in the top 70 of the 200+ ranking factors. Still only one part of a comprehensive strategy though, but important nonetheless.
That's not to say that it can't be gamed, much like links... but Google can use trust factors (Phone # verified accounts, IP address, diversity of IP location, multiple accounts, activity across those accounts, use of other Google services, search behavior, etc.) and influence factors (Followers to Followed ratios, Quality of Followers, Cross referencing links & social associations, etc.) that would help weed that out but probably never eliminate it.
I always enjoy the speculation, but we will see how it plays out.
Cheers,
Kevin
Absolutely Kevin. If you read some of the blogs about +1 you'd think that every other ranking factor was being thrown out. It's one more piece of a very very large pie.
That is what I am talking about, in some obscure nitches there wont be any evidence of social interaction in play, I mean are the producers of pencils going to ever have a natural followers? likes? +1 I dont think so.
I believe it safe to say that Google would have tested +1 even to the point I would be confident that they have an idea of what to expect as "natural occurrence" with +1 just like in other link profiles. A flood of a million +1 in a very short period of time would be very easy for Google to pick up on their radar as SPAM+1
Good point though :)
Yeah I know for sure they will have some time delay style thing. So if 1 million +1's come in for a hour you may rank for a good keyword for a few hours/days but then it will drop off, similar to the Tweets/retweets power ;)
Looking forward to this new feature - I can't see it yet.
Refering to the statement, "How will Google persuade regular people to create their Google profile page and add their friends?" --> the publisher button that can be embed on the webpage would certainly help. Then the non-Google profiles get in touch with that more often.
You can activate +1:
Thank you Gianluca, I have recognized it after my comment and tried out already :-)
Hi Petra. You can turn +1 on for yourself via Google Experimental here https://www.google.com/experimental/index.html if it is still unavailable for you.
Thank you Rob :-)
Wow. Totally awesome analysis.
And nice that Google start using social signals. SEO should become a more legitimate form of marketing with less manipulation. I hope.
This is a Fantastic Post, I feel it's heading down the road of people actually influencing search results. We can look at it Google competing with Facebook for the "like" space or Google empowering people to influence Search results more relevant to the query posted. We are already at the drawing boards. Either way, this is a great post and the data provided is brilliant.
Again one more filter by Google to get refined searches. Surely this will help user to check out most apporiate result matching to their queries and definately go for recommended searches.Its popularity will increase through networks of people over the ineternet.It can drive most ethical and quality organic traffice to the website.But at the same time,spammers can use this feature for their purpose.
This one of the important measure in social searches.Thanks for post.Keep updating.
Tom,
I +1'd SEOmoz blog... then used a little onmouseover, onmouseout, onclick to provide proof as well as an example for people that haven't joined the Xspearmint!
I think Social Signals are great, and am hopeful for +1 being among them. But, as everyone has pointed out, the possibility of abuse is there. I made a little proof of concept, which if you are +1 enabled and on Firefox or Chrome can embed the +1 button onto a site, and but it can be for another site completely.
Google need to come up with a good defence to this problem or it'll dilute the signal and we'll be back to square one.
"this is pretty cool" might a positive signal leading to a click through, but what if "you should check this out..." was intended to check out something 'bad'? then we're in trouble...
As long as Google has a way to monitor all the +1, I think this is a great idea, but we must be cautious about others abusing it, as IT WILL influence page rankings
+1
for the post of course ;)
For me, this is another move to Google on this Social Search. I haven’t checked it yet, but sooner or later I will… The main reason why Google is doing this is because Facebook. I´m not a FB “super fan” person, but majority of people is. So, as you said: Google is smart. They are trying to follow the mass.
On a real perspective, maybe people won’t notice this “+1” option in a fist glance. I´m not saying this is going to be like the green page rank on their Toolbar, but in my opinion a regular Google user maybe will keep sharing “everything” on their FB profile. One more final thought: people go to “search” on Google instead “share”. So, if your network of friends doesn’t use it, maybe you won’t.
On the other hand: What about the spammers? You can get thousands of Gmail accounts for few dollars in the “Gmail Bank”. This reminds me the Google Side-Wiki and others...
FANTASTIC post! It's a bit of a cliffhanger though - I'm dying to know more about the correlational data between social & traditional signals!
The announcement of +1 is huge. This is another real validation of the importance of social signals to search. I can imagine a room full of engineers looking at similar data showing correlation between likes and linking root domains... Google knows that it needs to own the "like" graph, and this is how they are going to get it. The visible social product is great too - non-intrusive, and really makes sense coming from Google.
The like button (and soon the +1 button) are beautiful ideas - it's a very low friction way for people to curate the web as they browse, and the resulting data is probably better than the link graph in many ways. It's much deeper (I see pages with no visible links but lots of "likes" all the time), more timely, harder to game (I'm convinced - WAY harder to game than links), and has so much more inherent meaning than a link does.
Will we have a total content meritocracy overnight? No, but +1 speaks volumes about how Google intends to move us toward one.
It's a great trying to build social SEO through Google. But the "+1" will be voted before knowing the content? I guess it will be used just to hack the results by people who knows it. Maybe Google should open a pop-up within the result so you can check the page and then give your "+1" vote.
If a bunch of SEOs get together in a private group to +1 each other's sites and clients' sites, how long do you think it would take Google to spot that algorithmically? Link farms worked for a while...
I know that great content is extremely important but I have a fear that new websites, even with great content, are going to find it even harder to do well in the search engines. As someone who is involved with launching new brand sites and a recent ecom site I see this as yet another hurdle. Anyone have any thoughts on how this will effect new websites?
Are we sure this isn't Google's April Fools Day joke?
After seeing all the Google April Fools jokes it makes me wonder how much money/man hours were lost to it.
After seeing all the Google April Fools jokes it makes me wonder how much money/man hours were lost to it. -Doh! Hit enter twice and it won't let me remove it.
The thing I like most about the +1 button is that it adds a potentially new user story to the mix of curating content.
Before, we could only perform and refine searches to find what we're looking for (or arrive via social network). Only after finding the great content could I voice my opinion on it via a tweet or like.
But with +1, we now have the option to add commentary before we even get to the content we want to see. Imagine, performing a search and seeing content from one of your trusted sources (e.g., SEOmoz) sitting below the result from a content farm. I believe many people will begin to use +1 as a way to correct what they see in the search results.
Not only will +1 be great on actual web pages, but I believe it's a game changer for the SERPs as well.
I'm glad to see so many other people are concerned about how this could be used by 'fake' people to promote a site; that was my first fear. I agree that Google knows enough scary data about each of us to decide if we're real or not.
Now that I think about this more, I think it will really put pressure on the competition for all of the 'review' and 'testimonials' sites..which is a good thing(?) It will be interesting to see how they respond.
I think this is the next step of web evolution, you could think spammers may manipulate this, but as Tom says "Think about how much information Google has on you", I think they are smart enough to discount value from +1s from accounts that are totally new and with no use at all. For me this is moving forward a web created and maintain mostly by regular people more than webmasters, where rankings are based on people opinions world wide rather than incoming links (which can be manipulated).
I'm just glad that there isn't a -1 so that competition can sabotage a given reputation like they can by leaving bad reviews on Google places. This would be abused far more than the +1 if -1 existed!
I can't count the number of times I've seen a comment on a Facebook status - when someone is ill, or has borken a bone, ornot be offered a job they wanted - along the lines of "Big unlike!" Having an Unlike button in Facebook and a -1 button in Facebook would a very good thing if the intention is to convey a sentiment to your social group. In fact, the only time it's a bad thing is if this "social" interaction's promary intention is actually to affect ranking changes.
This is pretty huge. Cue some of the big players leasing a few thousand mobile lines then, eh?;)
Little brainstorm I just did after reading this awesome blog post simply by asking "What if":
As a UK website, I wonder if this is another US centric social media add-on that won't really take off over here?
I’ve opted in to the experiment and it doesn’t work for me yet. Interesting they are so obvious in their copying of the Like button. Not exactly subtle Google.
Think the issue for us is whether UK people even browse logged in. Google have had the issue in USA that people aren’t always logged into profiles when browsing, not sure how we would push to our customer base when we have no idea if they’re signed in, as it’s not been such a big deal in UK thus far. Closest I can think is contacting customers we know are gmail users with some incentive to +1.
Hiya mate,
Makre sure you aren't getting redirected to Google.co.uk. Try going to Google..com/ig and running the search from there, which should show you the buttons now. :)
-Tom
Thanks for that, I'll try it now
I think this is a very interesting move by Google, so I'm dedicating a +1 blog to the subject. Will try to do some cool experiments soon. Anyone got any ideas what to test first?
Try a headline test with similar body content, with one headline aiming to be "viral material" and one a "normal" headline. See what it does for your +1's. I would also like to see the difference between static pages (normal URLs) and dynamic ones (blog posts with comments). Just some suggestions that I'll be testing myself :-)
Well almost 6 years later and the +1's are still going strong. Turned out not as big of an SEO impact as first thought but still important none the less.
I think that this is a really exciting development - it'll be interesting to see how it plays out, from an SEO perspective.
Over here in the UK, Digg is not massive or mainstream, Twitter and FB Likes seem to be the main area where I can see this gaining traction.
When I first read about the +1 I thought it is Google's April Fool's joke, so I came instantly to SEOmoz to see if there is something here about the topic but couldn't find anything. A couple hours later here it is what I was looking for. Great job Tom.
If this is going to be really an SEO signal than we have another conversion goal to track and optimize for.
I just +1'd this message. I have good practice.
Great post. I printed it. VERY important. Thanks
+1 is a great concept; it will revive Google's growth for the next few years. If you have savings - buy GOOG shares and hold them for a year.
And don't forget to +1 me when you cash out :)