Now there’s one thing you can be certain of whenever there’s a new launch from Facebook there’s going to be a huge amount of speculation about what it means for digital marketers; but now they’ve made a stab at launching a search engine - there will be even more than usual.

First, let's get the biggest caveat out the way, I’ve not used the search functionality yet, I’m on the beta waiting list, along with thousands of others, but I have poured over all the write ups including this excellent write up of the story behind by Steven Levy, which I think even tops his great piece about the launch of Google Plus. 

Despite that lack of practical experience of the new functionality, for the last couple of years I’ve thrown myself into understanding the Edgerank Algorithm, aka the rules that decide what appears in your newsfeed and come from a search marketing background. 

I wanted to collect some observations and ideas that have already come to mind and hopefully trigger a discussion about how Graph Search might impact on SEO and digital marketing in general.

Affinity is more important than ever

One of the three pillars of the Edgerank algo is Affinity, it’s fairly safe to assume once Facebook finds a relevant result for your search query they’ve got to decide which order to show them in. The affinity that you have to the individual responsible or connected to that content has got to play a part in deciding that order.  

To quote my Econsultancy article on Affinity:

Affinity is a score based on the proximity to or how “friendly” you are with someone. You’ve probably seen this in action. Spy on an ex-boyfriend or girlfriend, snoop on their profile and suddenly they’re in your news feed all the time.

Comment on someone’s photos and you’ll find them appearing in your feed more often. This is affinity in action. You’ve sent a proactive signal that you have a ‘close-ness’ to that individual or organisation. The algorithm acknowledges this and begins to order the results in your newsfeed accordingly.

It’s not far fetched to assume that this would be part of the sorting process. 

But weight’s got to be part of it as well...

Weight is another part of the Edgerank Algorithm, which is based upon the idea that some pieces of content, especially those with greater interaction, ideally from those you have affinity towards, are more deserving of attention in the newsfeed.

Typically when talking about the newsfeed Weight is discussed in reference to the type of content, i.e. videos having more weight better than text etc.

However, with Graph you won’t have different types of content competing, if you search for photos you’ll only get photos, if you search for restaurants you’ll only get restaurants, but weight will probably still play a part. 

Say I search for Photo’s of my friend James in Brighton during 2012, something’s got to determine the order, it probably won’t be affinity because they’re all photos of James (though some might taken by people you have affinity towards) and weight would seem logical, i.e. photos with lots of likes, shares, comments etc. appear above those with less.  

In fact one way of thinking about Graph Search is less like a search engine but more like a super-powerful filter of the newsfeed of the past. 

Multi-dimensional searches 

One of the really interesting way’s in which Search Graph works is the way it allows multi-dimensional searches. Danny Sullivan sums it up well:  

Another difference is the layers of searching or refinement that Facebook Search offers compared to Google. For example, a Google search can show you restaurants in San Francisco, a pretty much single dimensional view.

A Facebook search can show you restaurants in San Francisco liked by your friends. Or further, those liked by your friends who actually live in San Francisco, as opposed to those who live elsewhere. Or those liked by your single friends, your straight friends, your gay friends, your friends who work for a particular company….

This is something that’s really powerful and something fans of structured data have talked about for years. But this is structured data that nobody else has access to. A clever, and fiendish play by FB.

This is a huge deal in so many verticals...

Looking to buy from a service industry business like a plumber, solicitor, photographer etc? You can find those recommended by your friends, but that’s only scratching the surface, this has the potential to disrupt the way people choose businesses or find information in areas as diverse as:

  • Dating.
  • Hotels.
  • Restaurants.
  • Shops.
  • Recruitment.
  • Travel Destinations.
  • Tickets & Music.

The bad news for Google is at the moment the only vertical I can’t see being potentially affected by Social Graph is financial services.

At the moment it doesn’t index Status Updates

One of the most ‘rich’ text content is status updates and notes, at the moment Search Graph doesn’t index them or notes as they feel it’s too intense a process, this is interesting for a couple of reasons firstly if they decide to turn this on it’ll become even more powerful and secondly it means that currently the search engine is about entities rather than documents.

Entities are a huge part of how Google thinks, and talks about, its Knowledge Graph, if you’ve not come across this idea before you should read this article from Google.

Graph Search not expected on mobile anytime soon 

At the moment Search Graph isn’t going to be available on mobile but once it does there’s going to be one of the most powerful search moderators in history available ‘near me’.

It’ll be simple to find coffee shops my friends have like or checked into which are near-by. Again hugely powerful stuff. It’s something lots of start-ups have tried to nail but they’ve never had the richness of data and the social graph available to Facebook. 

Expect more from Bing 

Bing have been Facebooks search partner for sometime, they're still involved as you can read about here, but I bet they've had some interesting meetings recently. They're still powering Facebook's web search but it'll be interesting to see how that plays out in the coming weeks and months.

So what do businesses do? 

Well a lot of this is built around structured data so it’s going to be important for businesses to make sure they’ve got all the data they could provide about their business in the right place and in a format Facebook like.

Then they’ve got to tick the boxes of Edgerank, so when they do appear in a search result they appear as high as possible.

That means doing things like structuring campaigns to be constantly building affinity, especially with people who are themselves widely networked and have achieved a high affinity with they’re network.

Also understanding the peaks in search traffic and trying to make sure affinity is at an all time high at those points in time. Also ‘entities’ which you want to appear highly in results pages need to accumulate weight by having high levels of interaction.

It’s going to take a while to shake out, but it’s exciting and I think a genuinely different and compelling search experience which I think will put a lot of start-ups out of business and potentially put more pressure on Google than they’ve had in recent years.

Kelvin Newman

Published 16 January, 2013 by Kelvin Newman

Kelvin Newman is SiteVisibility's Creative Director and is the editor of the UK's most listened to Marketing Podcast. He also spends his time at conferences, tweeting too much and working on top secret research and development projects. He's also on Google+

21 more posts from this author

You might be interested in

Comments (6)

Save or Cancel
Kelvin Newman

Kelvin Newman, Creative Director at SiteVisibility

One thing I missed that's really important at this stage is seems things like photos from brands aren't being tiggered h/t to Thomas Baekdal.

Seems a bit of an oversight to me, If somebody likes a brand and their query could trigger a brands content it should appear. At the moment seems for businesses to be triggered it has to be a bit more explicit i.e. dentists my friends like

over 5 years ago

Ryan Sommer

Ryan Sommer, Freelance Consultant

Thanks for the insightful read Kelvin. Always enjoy your posts here. I am particularly excited about Graph Search when it comes to the first four bullets in your verticals impacted area above.

Coming from SF, and knowing how "gamed" Yelp is, I never understood the value of reading an online review from a stranger.

I've been a fan and member of the Zuck army since 08, and I know this is about to change things for the better, as long as MS doesn't muck it up!

over 5 years ago

Mark Gavalda

Mark Gavalda, CEO at Mandloys Digital Agency

Yes, the idea of social search (especially restrained to your friends or maybe one circle above) seems to go in the right direction however after the sponsored stories move of Facebook I'm already afraid they'll sacrifice the integrity of this service to get a few more dollars out of brands. Think sponsored reviews or maybe your Café won't even appear in the search results _unless_ you pay Fb a chunk of money :-/ Let's hope I'm wrong.

over 5 years ago



Thanks for sharing your views. Great blog here.. It’s hard to find quality writing like yours these days. I really appreciate people like you. I would like to thank for the efforts you have put in writing this blog.

over 5 years ago


Matt Isaacs, Senior Ecommerce Manager at Perricone MD

The actual technology looks amazing. Multi-dimensional search? Great!

However, the crux here is what is a geniune 'like'? I have had endless sponsored links come up for people who have a liked a brand on the basis of getting freebie or some such and then just immediately hidden them from their newsfeed.

The problem with Facebook likes, or any other social ranking system is always the question of "why would anybody publicly 'like' something anyway"? Why would I want to delare this and make it public and permanent?

At least with a customer review you have a segment of consumer who will genuinely enjoy sharing an opinon so there is some integrity there.

I much prefered the old system of toptable that rewarded points for reviews. You were in no way expected to leave a good review and the fact that you had to write something (as opposed to just a star rating) meant that most would engage and write something credible.

Again easily gamed to some degree, but with a good mix of real reviews in there too.

over 5 years ago


jeremy speiser 

Kind information

over 5 years ago

Save or Cancel

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 Digital Pulse newsletter. You will 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.