We all know it’s coming. Actually, many of us in digital marketing are surprised it’s taken so long. I’m referring to a real Facebook search engine, which has the potential to turn the search industry upside down. Mark Zuckerberg explained at Disrupt a few weeks ago that Facebook will tackle search, and that they are “uniquely positioned to answer the questions people have.” He also explained that Facebook sees about one billion searches a day, and that’s without even trying. Last, and most importantly, he said, “At some point we’ll do it. We have a team working on it.”
Facebook’s COO, Sheryl Sandberg, also commented about Facebook search recently, explaining how the recommendations of your friends can be important for users looking for information. She spoke about how the social network can tap into the “wisdom of friends” over the “wisdom of crowds” in a recent interview with CNBC. Needless to say, with both Mark and Sheryl talking so publicly about search, you know it’s only a matter of time.
Debuting at Number 2 and the Revenue Implications
One thing is for sure, the potential revenue implications for Facebook are huge. And that’s especially important as Facebook struggles to find sustainable sources of revenue. Search can be a powerful way to drive more revenue (if you can provide strong results, gain users, build query volume, and then monetize those results with relevant advertising).
From a user perspective, Facebook has over one billion active monthly users. From a query volume standpoint, Facebook already has one billion searches per day. There’s quite a bit of ad inventory already.
Think about this for a minute … Facebook has so much query volume now (one billion searches a day) that it would turn into the number two search engine overnight (behind Google). Google sees approximately 100 billion searches per month globally, and if Mark’s statistics are correct, then Facebook would come out of the gates at approximately 30 billion searches per month. And by the way, that’s with horrible search functionality in place now! I often joke that you can’t find yourself via Facebook search. 🙂
Facebook sees one billion queries per day, and without even trying:
New Engine, New Ranking Factors
So what if Facebook launches a killer search engine? What will it look like and how will it work? And more importantly for digital marketers, what will the ranking factors be?
As Mark and Sheryl both said during their interviews, Facebook has the ability to handle search in a different way than Google or Bing. They believe the power of your friends could drive the answers you require. I started to think about their quotes as soon as I heard them, and it was easy to see that Facebook search can definitely get interesting for webmasters, SEO’s, Social Media Marketers, and Paid Search Marketers.
But, until Facebook launches a serious search engine, SEOs will continue executing, analyzing, and refining their efforts based on a number of SEO ranking factors related to websites, webpages, links, on-page optimization, domain authority, etc.
Sure, Google and Bing continually update their algorithms, but we know that on-page optimization is important, building high-quality links is important, pumping out high-quality and unique content is important, etc. So, what happens if a big player jumps in the game with a completely different set of ranking factors? And I’m not talking about a slightly different set, I mean really different.
Facebook search ranking factors could be very different from traditional SEO ranking factors:
SEO’s and Social Media Marketers
Depending on the path Facebook chooses to travel with its search algorithm, SEO’s might find themselves neck deep in social activity (or trying to get neck deep in that activity), while social media marketers might hit the ground running and leverage their already strong social skillset. As many SEO’s already know, SEO and Social have been blending more and more over the past few years. To me, it’s hard to execute a strong SEO strategy without Social Media Marketing playing a role.
But, there definitely are many SEO’s and webmasters that haven’t taken the social plunge (and least not full blast). For those people, a Facebook search engine could lead to a very uncomfortable experience. That’s because the ranking factors could be so engagement-focused, that many of the traditional SEO factors might have very little effect on Facebook search rankings. Talk about a game-changer.
The Power of Your Friends or The Power of Similar People?
Before we dig into the potential ranking factors, I wanted to mention an important distinction. There are some people that believe social search should reveal answers to your questions directly from your group of friends. I believe that can be powerful, but only if that tight group of people truly holds the answers … With most people only having 100-200 friends on Facebook, that’s not realistic (in my opinion). And that’s where Google shines.
Ask Google any question, no matter how complex, and you’ll find an answer instantly. Even if that person with the answer is located half way across the world, and has no social connection to you. But wait, they actually do have a connection to you—a similar interest.
Even if you never met that person, they might have a similar interest to you, which means they have similar questions about that interest. That positions them in a strong way to answer questions you have about that topic, and vice versa. And this is what I believe will be at the heart of Facebook search.
If Facebook could effectively connect you with friends, friends of friends, and other people with similar interests on the largest social network in the world, then Facebook search could indeed be a serious player in the industry. And if those results provide amazing answers to your questions, then the revenue could start to flood the social network.
Google generates approximately $40 billion in revenue per year, and 96 percent of that is from paid search. So yes, Facebook could get a serious shot in the arm from its upcoming search engine. Actually, it could transform its business.
Friends can influence search rankings, but so could friends of friends, and people with similar interests:
Potential Facebook Ranking Factors
Facebook sees an incredible amount of data on a daily basis. With now over a billion active users, there are tens of billions of updates, shares, likes, etc. occurring on a regular basis.
Facebook knows what is getting shared, by whom, and where. It understands what’s getting hot, what’s being shared quickly, and what’s going viral by country, region, and city. It understands age, gender, interests, hobbies, the industry you work in, and more. It knows if you’re single, married, actively dating, divorced, etc.
The core point is that Facebook has a darn good profile of you, and your friends, which can help it determine and craft an interesting search engine results page (SERP) based on your queries.
Also, for the purposes of referencing the Facebook search algorithm throughout the rest of the post, I decided to name it BeastRank (named after Mark’s dog). Sure, I could have named it EngagementRank, SocialRank, FaceRank, or ZuckRank, but I have a feeling Beast could brand the algo. 🙂
So, without further ado, let’s take a closer look at possible Facebook Search Ranking Factors:
1. Pure # of Likes for a Piece of Content
Let’s start with a simple metric. Facebook knows every single like that occurs across the Web and on Facebook. Facebook can also see the velocity of likes, including how likes are being sustained over a period of time. For example, is the content receiving likes over a long period time (evergreen content), or is it going viral for a short period of time and then tailing off?
For SEOs, likes could be the equivalent of the raw number of links pointing to a webpage. But like inbound links, all likes are not created equal … I cover more about that soon.
Likes per piece of content could be similar to raw links per webpage:
2. Pure # of Likes for a Given Domain
So, if “likes” are similar to links, then “likes by domain” can contribute to domain authority. Facebook can understand how many likes a given domain or subdomain has gotten over time. It can understand if a site is new or established. It can determine if it is receiving likes from dummy accounts, new users, or trusted accounts.
Likes by Domain can absolutely help Facebook tailor the right search results by weeding out low-quality domains based on their history. Also, Facebook can determine “likes by domain by category” to start to understand the power of a given domain within a certain niche. More about categories and interests soon.
3. User Engagement for a Specific Object Shared on Facebook
Facebook already has Page Post metrics for “Engaged Users,” “People Talking About This,” and “Virality,” and you can view these metrics in your Page Insights. Facebook could easily use these metrics for all types of content shared on the social network, which could influence the search rankings.
For example, how many users clicked on a specific object (engaged users), how many users generated new shares for piece of content (people talking about this), and how many users engaged a piece of content based on the total number of users that saw that content (virality).
In order to calculate these metrics for all content shared on Facebook, and not just Page Posts, Facebook would need to analyze the destination URL for objects shared on the social network. For example, understanding when an external piece of content is shared across thousands of users, what the engagement was based on those shares, etc.
By the way, can you see why Open Graph apps are important, and how they can impact BeastRank?
Facebook could leverage its page post metrics to score all content shared on the social network:
4. Engagement within a specific country, region, or city (Likes, Shares, and Comments)
Depending on the type of query, understanding location is important. Facebook can absolutely understand engagement by geographic location based on a number of factors, including your IP address, the location you include in your profile, GPS (for mobile), check-ins, etc. This can help Facebook determine the right search listings by country, region, city, town, etc. This could also help drive Facebook Local Search.
In addition, Facebook can determine how many likes, shares, and comments a local business has gotten from users in that geographic area. For example, which steakhouse should be returned in the search results when I’m in Princeton vs. New York vs. Philadelphia? Facebook can use engagement by location to influence the rankings.
Understanding engagement by location could help Facebook tailor local search results:
5. Engagement within a specific category, industry, or niche (Likes, Shares, and Comments)
If a brand page, status update, or webpage gets 10,000 likes, should that page automatically rank higher than another that received just 1,000 likes? Similar to Google, the answer is absolutely not! It depends on topic, category, freshness, relevancy, etc. Facebook can understand the universe of people liking a specific object and then determine which types of interests and categories that object should rank for.
Let’s face it, there are times that a page ranks well in Google without having mind-blowing SEO metrics. And there are times those pages rank above other pages located on very powerful domains. Understanding interest and category can help Facebook reveal content from lesser-known domains that are extremely relevant to a search query.
Understanding engagement by interest can help Facebook tailor search results per user:
6. Engagement within your group of friends.
Both Mark and Sheryl have mentioned this in the past, and I agree to an extent. If your friends on Facebook can help answer a question based on recent updates, likes, shares, etc., then that content should be presented prominently in Facebook search. So, a page that’s been liked or shared by a friend could be given more weight in BeastRank, since Facebook wants to provide “as personal a search result” as possible (as long as that result is high quality and could answer your question). I mentioned this earlier in the post.
I also believe there are several dimensions to this factor. For example, Facebook could weigh in certain actions from your friends more than others. Maybe an initial update from your friend has more weight than a like. Maybe shares and comments on a specific piece of content mean that person is more interested in that topic.
And of course, Facebook will be able to turn the levers up and down, based on user activity, engagement, the quality of search results, etc. (just like Google does with Search Plus Your World).
The “wisdom of friends” vs. the “wisdom of crowds”:
7. Engagement within friends of friends.
Expanding on the bullet above, friends of your friends could hold answers that you are looking for, too. Hey, they might have more in common with you than you think. For many people, friends of friends have similar interests, locations, age, etc. I’m not saying this will be the most powerful signal Facebook could use, but it absolutely could be included in BeastRank.
For example, I’m friends with some people in the SEO industry. If they have connections that are also involved with SEO, and I’m not friends with them already, then maybe their social activity could end up influencing my search results.
8. Engagement with Content (Click-Through and Dwell Time by Demographic, Location, and Interest)
If Facebook knows how many times a piece of content has been shared, then it also understands click-through rate of its users (and by segment). If Facebook sees consistently-high click-through for a piece of content shared on the social network, then it could place a score on that content.
On the flip side, if Facebook sees very low click-through, then maybe there’s a reason users are not engaging with the update. That could negatively affect the content in BeastRank.
Also, Facebook could determine how users engage with certain types of media content (like videos and photos). For example, if a video is shared, then did they start the video, how long did users watch, how far did they get through the video, and what was their dwell time?
If a photo is shared, did users click the image and view the photo page, and what was their dwell time? These signals could influence rankings of blended search elements like photos and video.
Understanding click-through, dwell time, and media engagement could provide signals to Facebook:
9. Brand Page Influence (Engagement by Brand Page)
If the content being shared can be tied to a specific brand page, then likes and engagement by brand page could influence the rankings. This is another variation of domain authority mentioned earlier, but based on a Facebook page versus external domain. Facebook could determine when the page was created, how many likes it has built, how quickly those likes were built, engagement by its fan base, virality of page posts, location of engaged fans, interests of engaged fans, etc.
In addition, you can check out Page Insights to learn more about “People Talking About this Post,” “Virality,” “Engaged Users,” which are metrics tracked for each page post. Since Facebook already tracks these metrics, it could score a brand page based on the history of page posts.
If Facebook flags a brand page as high quality (over time), then those signals could help that brand page’s content rank in search. And on the flip side, it could help Facebook bury that page’s shares and updates. Think about Panda for Facebook Pages. 🙂
Building Brand Page Authority could be the equivalent of Domain Authority:
10. On-Page, I mean On-Post, Optimization
Yes, a traditional SEO ranking factor makes an appearance! Facebook could analyze the post, update, or share, along with the destination page to determine what the content is about (theme, keywords, category, etc.) This can obviously help Facebook determine a certain level of relevancy for a query. For page posts, status updates, and shares, Facebook could analyze the body of the update, the title and description of the content being shared, etc.
Traditional Content Optimization Could Very Well Play a Role in Facebook Search Rankings:
Also, Facebook could analyze any comments that users are adding for a given piece of content (in Facebook and off-site). All of these elements could help Facebook better understand the topic of the content, the categories and keywords that match up, etc.
Note: External crawling and document analysis is no easy task, and could lead to a lot of spamming when Facebook search launches. Let’s face it, we all take Google for granted … That said, when combined with the social search factors I’ve been listing, you would hope that BeastRank could determine high- and low-quality pages and domains and adjust the rankings. I’m hoping that’s the case, or this will be the Wild West on steroids. 🙂
11. Page and Domain metrics via Moz or Majestic (huge opportunity for these companies…)
Now, Facebook might tell you that social can reveal better rankings than traditional search, but I believe there are traditional search factors that can be extremely valuable to Facebook’s upcoming algorithm. Both SEOMoz and Majestic have proprietary metrics that take into account a host of SEO factors.
For example, mozRank, mozTrust, Domain Authority, Page Authority, Trust Flow, Citation Flow, etc. It would be very smart for Facebook to tap into these metrics when determining rankings. They can absolutely help Facebook understand more about the external factors for a given page or domain.
Third party SEO analytics can help Facebook understand off-network statistics:
12. Influencer Engagement (overall, and within specific categories).
I hope you didn’t think I would write a post about social search without including influencers! Every update on Facebook, like, share, comment, etc. is not equal, just like every link to your website is not equal.
If someone with 1,200 friends and 45,000 likes on their brand page shares a certain piece of content, then Facebook could weigh that user higher than the person that just joined the social network, without a strong history, and without a lot of connections.
All likes, shares, and comments are not created equal. Influencers will play a role in weighting content:
In addition, Facebook could easily analyze its industry, niche, category, etc. to determine its focus. For example, I might influence someone interested in digital marketing, the New York Yankees, and horror movies, while I shouldn’t influence anyone interested in baking, the New York Mets, or fried food. I fully expect influencer scores to impact rankings in BeastRank.
Note: Is this how Klout could find a home? Influencer scores are hard to accurately report, and Klout has been on the forefront of this category for some time. I don’t think Klout works perfectly by itself, but when combined with all the other data Facebook has access to, then it could help influence BeastRank. It’s worth noting that Klout is working with Bing (Microsoft) on its social sidebar. And Microsoft was an early investor in Facebook. It’s not hard to see the connection… 🙂
Final Thoughts and Next Steps:
If Facebook Search comes to fruition, then the importance of Social Media Marketing will increase exponentially. The game will have changed—and in a big way. Companies leveraging their SEO power to gain a significant portion of their traffic, revenue, and profit will have to look at Social now. That’s unless you want to avoid the number two search engine in the world.
Below I’ve listed several bullets explaining what you can do now to prepare for the upcoming Facebook search engine. And by the way, this will help you now, even before the engine arrives:
- If you haven’t already, then start engaging in social media marketing now. If you don’t form a strong social strategy, you could absolutely be left in the dust when Facebook search hits.
- Enhance your content generation plan. Do not sit on current SEO rankings and think you’re in great shape. In order for Social to work for your company, you need great content. This can take many forms, so do your homework, understand your industry, your customers, your prospective customers, etc. Create a killer content generation strategy knowing it can help you answer questions, impress others, increase shares, exposure, links, and rankings.
- Integrate Facebook social plugins on your site. Millions of sites have done this to some extent, but I would look to enhance that connection now. You should also start thinking about Open Graph apps.
- Leverage Facebook’s platform to its fullest to enhance your current presence. For example, take advantage of photos, video, events, etc. In addition, utilize Facebook’s advertising platform to increase your exposure, reach more people, and gain more likes. Take a serious look at your fan base, and look to enhance that number and how you connect with them. This can only help you when Facebook search hits (as you can start with a stronger base than you have now).
Summary: Are You Ready for BeastRank?
As you can see, Facebook Search ranking factors could indeed be much different than traditional SEO factors. And that can have a huge impact on how digital marketers work, especially how it relates to SEO, social media marketing, and paid search.
But the proof is in the pudding. It’s easy to say you’ll create a killer search engine when you don’t have one. Understanding data at scale and providing an incredible search experience is tough stuff. We all take Google for granted, and that becomes apparent when a new “Google Killer” hits the scene (remember Cuil?)
Needless to say, I’m eager to see Facebook roll out its much-anticipated search engine. It definitely has the potential to rock the search industry. And if it does, will you be ready to rock on or rock out? Start thinking about it now.
Subscribe to SEJ
Get our daily newsletter from SEJ's Founder Loren Baker about the latest news in the industry!