{{ searchResult.published_at | date:'d MMMM yyyy' }}

Loading ...
Loading ...

Enter a search term such as “mobile analytics” or browse our content using the filters above.

No_results

That’s not only a poor Scrabble score but we also couldn’t find any results matching “”.
Check your spelling or try broadening your search.

Logo_distressed

Sorry about this, there is a problem with our search at the moment.
Please try again later.

According to new data, Google +1s have a stronger correlation to high natural search rankings than Facebook Likes, and play a bigger role in a blended natural search strategy.

Digital agency Stickyeyes has looked at the influence of both Google +1s and Facebook Likes as emerging ranking signals using its Roadmap tool, which analyses over 100m ranking signal metrics each month across a 13 terabyte database.

This study looked at over 7,000 URLs that rank in the top 20 positions across multiple competitive sectors in Google.co.uk and correlated their actual +1 and Like counts versus ranking position. 

The results suggest that +1s have the strongest correlation with search rankings...

An ever-changing landscape

Google uses more than 200 core ranking signals to populate its search results and it is constantly refining its algorithm based on these signals.

This is demonstrated in the graph below, which is a follow on from a previous post which used Stickyeyes data, ‘The effects of the Google Panda update’.

(The graph shows movement of website rankings in Google. Higher peaks indicate greater position movement across a larger number of websites.)

As we are currently experiencing an unprecedented series of algorithm updates, understanding which signals have a stronger influence on rankings and then deploying a blended search strategy with these elements is essential. 

Google +1 vs. Facebook Likes

Stickyeyes looked at the influence of both Google +1s and Facebook Likes as emerging ranking signals by analysing over 7,000 URLs that rank in the top 20 positions across multiple competitive sectors in Google.co.uk and correlating their actual +1 and Like counts versus ranking position.

Explaining the graphs

 Each marker relates to a count (e.g. Google +1s) and indicates how many URLs out of the 360 URLs at each ranking position have this count or lower.

  • Diamond represents 50% of URLs.
  • Triangle represents 65% of URLs.
  • Square represents 80% of URLS.

The linking box indicates the spread of URLs between 50%-80%, selected as the ideal starting range for a URL wanting to achieve high rankings. A smaller box indicates a smaller spread of values across URLs. 

The grey plotted area spans the gaps between 50% and 80% of where URLs fall at each ranking position.

The smoother a plotted area with growth towards position #1 and less peak and trough’s between each position, the stronger the correlation to high rankings.

Google +1s

 

  • Google +1s have a rank correlation score of 22 out of 100 which indicates a positive trend. This is one of the higher ranking signal scores out of over 200 signals which Google use to decide position.

    Go far enough back and a signal such as Meta Description Keyword Density would have scored 80+ out of 100. The correlation score is designed to expose the weightings of unique signals within the algorithm and finding the right mix of the higher scoring signals is vital.

  • There is a clear upwards trend in the count of Google +1s per URL as we move from position 20 in Google.co.uk to position one.
  • The relatively short height of the linking boxes at each position show a closer scatter of +1 counts across URLs. The absence of any marked peaks and troughs and the steady upwards trend in the plotted grey area (with the exception of position 17) show a positive correlation to high rankings.
  • 23 or more Google +1s for your ranking URL would put you in the top 35% of ranking URLs, whilst 63 would put you in the top 20% making this an easier signal to gain traction in than Facebook Likes.

Facebook Likes

 

  • Facebook Likes have a rank correlation score of 16 out of 100. 28% less than Google +1s.
  • There is, however, still a positive trend of the count of Facebook Likes by URL as you move from position 20 to one, but the greater frequency of peaks and troughs indicate the correlation isn’t quite as strong as +1s.
  • The scatter of values between 80% and 50% is significantly larger than Google +1s. Whilst this isn’t surprising given that Facebook Likes have been available for longer and with larger uptake, it does mean it takes a greater effort to achieve the same relative counts.

    35 Facebook Likes on your ranking page or more are required to sit in the top 35% of URLs whilst 155 or more Facebook Likes are needed to sit in the top 20% of URLs.

How does it compare to other ranking signals? 

Whilst it is true that Google +1s have a stronger correlation to high rankings than Facebook Likes, it is important to note that one shouldn’t be actioned at the expense of the other, and that they must always be considered within the context of a wider search strategy consisting of a blended approach that hits multiple ranking signals.

Below is a typical example, showing that the combined authority of the external links going to a ranking URL has a positive correlation to high rankings only marginally lower than Google +1s.

Summary

Google +1’s have a stronger correlation with high rankings than Facebook Likes and are easier to gain traction on due to the lower relative counts per ranking position. These stats echo some of the findings from Kevin Gibbon's recent guest post

Graham Charlton

Published 23 July, 2012 by Graham Charlton

Graham Charlton is the former Editor-in-Chief at Econsultancy. Follow him on Twitter or connect via Linkedin or Google+

2565 more posts from this author

Comments (30)

Comment
No-profile-pic
Save or Cancel
Avatar-blank-50x50

Jenni

From the information you've provided here, this looks like an extremely flawed study...

about 4 years ago

Avatar-blank-50x50

John

What specifically makes you say that Jenni? It's quite the statement without suggesting why you feel that is the case.

about 4 years ago

Avatar-blank-50x50

Andrew Liddell, Ecommerce Business MGR at Personal

Its one sure way for Google to kick start their Social network.

@John, I thought the same!

about 4 years ago

Avatar-blank-50x50

Lyndon NA (The Autocrat)

There may be a correlation - but that is likely all it is.

As far as we know, as far as Google have said,
things like Facebook Likes and Google +1's have no direct influence in Ranking.

They may be utilised in the Social Signal bucket - used to vet link profiles and authenticate a bit of trust (poss. amongst other things),
but that is all they currently do.

The most likely cause for the correlation is that such sites rank higher (due to other means), and are marketed better.
Higher prominence usually results in more views ... and if you take a % of

Of course - that doesn't mean you shouldn't aim to obtain such things.
They hold CTR and CR influence.
They can help further perceived Trust to users/visitors.
And lets not forget - Google is working hard with the objective of using such signals (if they can separate signal from noise).

about 4 years ago

Avatar-blank-50x50

Andrew Liddell, Ecommerce Business MGR at Personal

Lyndon , the largest consortium of SEO specialists believe something quite different. The general trend is for social signals to be the fastest moving ranking factors in Googles algorithms, taking position 1 and 2 in the future of ranking factors. Currently contributing 7.2% at domain level and 5.3% at page level, keywords contribute 14% to ranking, why would anyone ignore social metrics? I don't know who you have been speaking to at Google, because thats not what they tell me!

about 4 years ago

Graham Charlton

Graham Charlton, Editor in Chief at ClickZ Global

@Lyndon I think, at the moment we can only point to a correlation, but with social signals likely to be more influential in search, it does seem likely that +1s will become a more important ranking factor.

Also, as Kevin Gibbons said in the post I linked to in the summary, "a strong social footprint is likely to make your rankings more defensible and future proof to algorithm updates".

This means having a strong content strategy and is something that, regardless of the correlation between +1s and SERPS, will have its own benefits.

about 4 years ago

Avatar-blank-50x50

Ben Lloyd

Love the disparity between Google +1s and Facebook likes on this article too ;)

about 4 years ago

Ian Brodie

Ian Brodie, Founder at The Rainmaker Academy

Was anything done in the study to factor out the rather blindingly obvious potential cause of the correlation:

Site that do better in the search engines get more visitors. More visitors = more people likely to click the +1 or Like button.

I would absolutely expect there to be a correlation between +1s/likes and search engine position - even if we knew for sure that there was no causal link. Simple because more visitors = more +1s/likes. It would be surprising if we saw anything different?

You just can't compare a correlation like this to a correlation with (say) meta tags or keyword density which is not impacted at all by having more visitors.

It's a bit like saying that because there's a lot of litter at football matches and there's more litter at matches with bigger crowds, then litter must be a cause of big crowds at football matches.

Unless you've factored it out somehow. Perhaps looked at sites over time to show that an unusually high number of +1s/likes today leads to an increase in rankings later?

Ian

about 4 years ago

Avatar-blank-50x50

Andrew Liddell, Ecommerce Business MGR at Personal

Ian, the study is not about total number of likes, but that Google+ 1s add more to page/site rankings than facebook Likes.

about 4 years ago

Keith Horwood

Keith Horwood, Head of Performance Marketing at CoinDesk Ltd

If anything, it makes sense that a Google +1 would affect natural rankings more than a like on Facebook, but of course that does not mean that Facebook can not drive significant revenue + traffic itself. You cannot really compare 'like' to 'like' in this case.

about 4 years ago

Ian Brodie

Ian Brodie, Founder at The Rainmaker Academy

Come again Andrew?

You're using correlation between likes/+1s to show that +1s have "more impact" than likes on search engine rankings.

My comment is implying that there may be a more obvious cause of the correlations - ie I'm saying that unless there's something you were doing in the study to remove the obvious cause of the correlation being that you get more +1s and likes when you get a lot of visitors (and you get more visitors when you're highly ranked) then your conclusion is meaningless.

I don't mean to go off on a rant, but there are some studies you can do to show whether +1s or likes have more/less impact on rankings - but a straightforward correlation isn't one of them - the variables aren't independent.

I could be doing your study a discredit. There could be more to the methodology than presented.

But right now it looks like it was done by someone who doesn't understand when the results of a correlation study are valid and when they're not.

Ian

about 4 years ago

Avatar-blank-50x50

kevin

Matt Cutts himself stated that Google does not use +!'s as a ranking factor, so what gives?

about 4 years ago

Peter Leatherland

Peter Leatherland, Online Sales Manager at Ethical Superstore

I think Ian is right, it is a correlation which is a long way off showing a cause/effect relationship, better search results means more traffic, and so more +1s and likes so which is causing which? It could also be a third unrelated factor causing a correlation for example;

Sites which rank well have put effort into SEO and probably have marketing resource, if they are putting effort into marketing and SEO it is likely the person/people responsible have suggested +1 and like buttons on the site, therefore the better ranking sites get more +1s and likes and the sites which have no SEO/marketing resource have not got round to putting +1 buttons or promoting on G+

about 4 years ago

Paul Huggett

Paul Huggett, Strategic Planning Director at Stickyeyes

Ian / Peter,

You make a really good point essentially asking if the signals are 'cause or effect' factors of ranking highly. They should always be considered in the wider context of the 200 core ranking signals Google uses and as with everything to do with Google's algorithm the data we’ve presented is very much open to interpretation – we’re certainly not saying that more Google +1’s will definitely result in higher rankings, the data however does allow us to infer that there is a likely relationship.

As Graham mentioned we can only point to correlation and as Lyndon suggests it may be that they are used to authenticate trust. It’s a little bit like the 'what came first - chicken or egg' scenario when you can't actually have one without the other. In this instance the only categorical way to prove/disprove would be a controlled URL case where the only metric that is adjusted is +1s - however this in itself is massively complex as you need to consider what volume to aim for, what deployment rates you use, which authorities you should try and engage with and then what all the other hundreds of signals should be set at.

about 4 years ago

Avatar-blank-50x50

kaminee

First of all, could you provide the basis of study done by Sticky eyes. Nothing of such sort could be found on their website.

about 4 years ago

Peter Leatherland

Peter Leatherland, Online Sales Manager at Ethical Superstore

@ Paul - I agree, with 200+ factors it is almost impossible to test each one fairly keeping the others all equal unless you set up websites specifically to test possible ranking factors. Because of this much is based on consensus

It is my opinion (and I can’t prove it!) that +1s don’t have a huge impact yet, I think Google is holding back with social signals as I don’t think they are as good an indicator as links. If Google was just starting up they would perhaps be a better measure to use than links, however Google has been using links as a significant ranking factor for a long time and they’ve done a huge amount of work (and spent a lot of money) developing an algorithm which can see around non genuine links and push websites to the top which have genuine high value links pointing towards them, it is much harder for a spammer these days to use spammy links to get top results.

I see likes/+1s in a similar light to links, had Google been fighting against spammers for the last ten years over fake likes/+1s they would have a robust system that can eliminate the effect of non genuine ones. They haven’t so I don’t think they can risk losing their status as the best search engine and plumping with social signals just yet. It’s not hard to go to fiverr and buy 1000 +1s

about 4 years ago

Ian Brodie

Ian Brodie, Founder at The Rainmaker Academy

Paul - you (or at least whoever wrote this article) has done an awful lot more than "infer there is a likely relationship".

The headline of this article is "What has the greater impact on organic search rankings? Facebook Likes or Google +1s?".

The article not only implies that +1s have an impact, it implies that the impact is greater than facebook likes.

It then goes on to talk about actioning +1s and likes as part of a search optimisation strategy.

That's an awful lot more than pointing out there's a relationship between two variables.

Correlation statistics are often misinterpreted - but this is a particularly bad example. Normally the correlation could be caused by a hidden third factor as Peter points out - and that's enough to cast doubt on the results.

But in this case there's a blindingly obvious cause for the correlation. Pages with better rankings get more visits. More visits = more likes/+1s.

By far the most obvious reason for the correlation is that high rankings are causing likes and +1s (because they get more traffic) not vice versa.

Now I'm not saying that +1s/likes have no impact on rankings. But your study doesn't show it.

You're also using the fact that +1s have a higher rank correlation than likes to imply that +1s have more impact on ranking (remember, the title of the post is "what has the greater impact...").

But higher rank correlation scores do not imply higher impact. They imply a greater likelihood of a relationship. Something with a lower correlation could still have a much bigger impact.

I hate to sound argumentative, but this is a perfect example of how unfounded rumours and myths about SEO start. Spurious statistics misinterpreted and repeated (just look at all the tweets with people going that one step further and saying "+1s have a bigger impact on rankings than likes").

Please either report the data correctly, or do the proper studies that show impact.

Ian

about 4 years ago

Graham Charlton

Graham Charlton, Editor in Chief at ClickZ Global

@Ian Paul created the data, I reported on it.

I posted this because, in the light of so many SEOs predicting that social signals would become more important as a ranking factor, I though this article would provide the basis for a useful discussion, and it certainly merits further investigation.

about 4 years ago

Avatar-blank-50x50

Andrew Liddell, Ecommerce Business MGR at Personal

@Ian, im not connected to this study in any way.

about 4 years ago

Ian Brodie

Ian Brodie, Founder at The Rainmaker Academy

@andrew - oops - sorry - I thought you were the study organiser or something.

@graham - so you're to blame ;)

It's true - it does merit further investigation - but I think the way the findings have been presented results in people taking this as a conclusion (ie +1 has the biggest impact, lets focus on that) rather than as a starting point for investigation.

about 4 years ago

Avatar-blank-50x50

Andrew Liddell, Ecommerce Business MGR at Personal

@Ian, thats ok.

I just see google pulling out all the stops in making sure like many other of their ventures don't spectacularly fail!

It would make sense for them to to add more 'weight' to +1s as they are in direct competition with Facebook on both advertising and search revenues.

You can already see +1s cues appearing on Google AdWords and businesses are rushing to make use of the first user advantage.

about 4 years ago

Ian Brodie

Ian Brodie, Founder at The Rainmaker Academy

@Andrew - that could be true. On the other hand if I were Zuckerberg and I heard that google was favouring +1s vs likes in their algorithm I'd have the lawyers looking at it...

What I'm sure is taken into account is the traffic that comes from social networks - ie the number of people who click to the page from google plus or facebook. And, of course, the more +1s and likes, the more that traffic is likely to be.

But it would be a tricky line for google to favour +1s in rankings I suspect.

Ian

about 4 years ago

Graham Charlton

Graham Charlton, Editor in Chief at ClickZ Global

@ Ian Yes, I'm to blame. As Editor, I also have to consider what has the biggest impact on my page views;)

I think the opening paragraph does mention quite clearly that this study is about correlation. I trust that the highly intelligent readers of this blog can look at the stats and draw their own conclusions.

I do wonder (and hope, since I've done this for the Econ team) if the authorship program may be used as a ranking factor in future. I would like to see more research into this.

It certainly has the potential to increase CTR, as these results are more prominent, buit will Google give higher rankings to authors that have joined Google+?

about 4 years ago

Andreas Pouros

Andreas Pouros, Co-founder & COO at Greenlight

Ian Brodie has this spot on. 13 years ago I drew up my list of rules for every new SEO that joined my team and top of that list was:

RULE 1: Correlation does not imply causality

Anyone conducting research needs to measure everything they do against that principle. Basic, basic stuff. </rant>

Graham, your last point about authorship as a ranking factor is very interesting. If you look over the half-dozen or so patents Google has filed in that area (like Agent Rank for instance) the suggestion there would be that authorship data would allow them to calculate a reputational score for an author, which can then be attributed to content they have produced, which is then re-scored based on the reputation score of those commenting on it. This would then potentially be used to determine which authors are 'hubs', which are 'authorities', build an old-school HITS vector graph for people, and at least weight links between content with that insight.

This however is a far more complex process than it sounds as it doesn't take into account the sentiment of the comments attached to a piece of content and, more importantly, not enough people have Google identities so not enough content out there is linked to enough people for Google to build robust vectors around.

Given this challenge, it's no surprise that Google is spending a huge amount of time and effort accumulating people's identities. For example, I've just bought a Google Nexus 7, which they are selling at practically cost (a smart move), it is preinstalled with Google+ and everything else Google, but you can't use a tenth of the device without logging into it with your Google details. There's also a rumour (I think it's still a rumour!) that Google are planning to convert all Youtube logins to global Google logins - given that Youtube is the 2nd most popular search engine, that's a lot of identities. My point here is that until Google reaches critical mass with IDs, I don't think they'll be factoring authorship into rankings to a strong enough degree for us to be able to measure it. That said, we'll try and run some tests - who knows, we may find that content written by people with the letter 'J' in their name or people that like cricket rank better than everyone else :-)

about 4 years ago

Avatar-blank-50x50

Stuart P Turner

Graham; just remove this line:

"The results suggest that +1s have the biggest impact on search rankings..."

Problem solved.

Interesting study and subsequent discussion...

about 4 years ago

Avatar-blank-50x50

andreas.wpvwpv

Great insight, I'd love to see the data behind it (perhaps not every record,though:-)).

How do you calculate the 'combined link authority score'?

Is is the average of the product of number of links with authority score from... seomoz/majestic/internal?

about 4 years ago

Avatar-blank-50x50

Sergiu Draganus

Interesting is to see how the whole SM is affecting the SERP rankings. I would like to see a similar case study about local reviews, as since Google Venice local google results are seen in the organic results directly

about 4 years ago

Avatar-blank-50x50

Santiago

This post will assist the internet users for building up new webpage or even a weblog from start to end.

about 4 years ago

Avatar-blank-50x50

Kunkel

Such a refreshing change to find good content for once, I am getting sick of the constant
drivel I find daily, thanks.

about 4 years ago

Avatar-blank-50x50

Oliver

A really interesting read, I started working with Google+ a little while back and concentrated on getting +1's through communities and it seems to be helping my search rankings.

over 3 years ago

Comment
No-profile-pic
Save or Cancel
Daily_pulse_signup_wide

Enjoying this article?

Get more just like this, delivered to your inbox.

Keep up to date with the latest analysis, inspiration and learning from the Econsultancy blog with our free Daily Pulse newsletter. Each weekday, you ll receive a hand-picked digest of the latest and greatest articles, as well as snippets of new market data, best practice guides and trends research.