(image) Dialing in from the department of Pure Speculation…
As we all attempt to digest the implications of last week’s Google+ integration, I’ve also be thinking about Facebook’s next moves. There’s been plenty of speculation in the past that Facebook might compete with Google directly – by creating a full web search engine. After all, with the Open Graph and in particular, all those Like buttons, Facebook is getting a pretty good proxy of pages across the web, and indexing those pages in some way might prove pretty useful.
But I don’t think Facebook will create a search engine, at least not in the way we think about search today. For “traditional” web search, Facebook can lean on its partner Microsoft, which has a very good product in Bing. I find it more interesting to think about what “search problem” Facebook might solve in the future that Google simply can’t.
And that problem could be the very same problem (or opportunity) that Google can’t currently solve for, the very same problem that drove Google to integrate Google+ into its main search index: that of personalized search.
As I wrote over the past week, I believe the dominant search paradigm – that of crawling a free and open web, then displaying the best results for any particular query – has been broken by the rise of Facebook on the one hand, and the app economy on the other. Both of these developments are driven by personalization – the rise of “social.”
Both Facebook and the app economy are invisible to Google’s crawlers. To be fair, there are billions of Facebook pages in Google’s index, but it’s near impossible to “organize them and make them universally available” without Facebook’s secret sauce (its social graph and related logged in data). This is what those 2009 negotiations broke down over, after all.
The app economy, on the other hand, is just plain invisible to anyone. Sure, you can go to one of ten or so app stores and search for apps to use, but you sure can’t search apps the way you search, say, a web site. Why? First, the use case of apps, for the most part, is entirely personal, so apps have not been built to be “searchable.” I find this extremely frustrating, because why wouldn’t I want to “Google” the hundreds of rides and runs I’ve logged on my GPS app, as one example?
Secondly, the app economy is invisible to Google because data use policies of the dominant app universe – Apple – make it nearly impossible to create a navigable link economy between apps, so developers simply don’t do it. And as we all know, without a navigable link economy, “traditional” search breaks down.
Now, this link economy may well be rebuilt in a way that can be crawled, through up and coming standards like HTML5 and Telehash. But it’s going to take a lot of time for the app world to migrate to these standards, and I don’t know that open standards like these will necessarily win. Not when there’s a platform that already exists that can tie them together.
What platform is that, you might ask? Why, Facebook, of course.
Stick with me here. Imagine a world where the majority of app builders integrate with Facebook’s Open Graph, instrumenting your personal data through Facebook such that your data becomes searchable. (If you think that’s crazy, remember how most major companies and app services have already fallen all over themselves to leverage Open Graph). Then, all that data is hoovered into Facebook’s “search index”, and integrated with your personal social graph. Facebook then builds an interface to all you app data, add in your Facebook social graph data, and then perhaps tosses in a side of Bing so you can have the whole web as a backdrop, should you care to.
Voila – you’ve got yourself a truly personalized new kind of search engine. A Facebook search engine, one that searches your world, apps, Facebook and all.
Strangers things will probably happen. What do you think?
Update: Facebook’s getting one step closer this week…
In many ways I think Facebook is already a search engine, it’s just not the type of search we’ve grown accustomed to over the last decade.
http://www.blindfiveyearold.com/facebook-launches-search
But even if you think the current social curation of relevant information isn’t a “push” form of search I think Facebook could quickly evolve and deliver search results to users.
http://www.blindfiveyearold.com/the-future-of-facebook-search
The Open Graph and Facebook’s topical modeling should allow them to surface related items based on what’s showing up in my News Feed. This type of “assisted search” would move intent further from passive and closer to active, which is what I believe most of us think of as traditional search.
Facebook has us busily creating a curated version of the Internet and it seems like only a matter of time before they truly leverage that asset. Then again, I figured they would have done this already, so perhaps Facebook has a different vision.
Google > Facebook…
Facebook is already a search engine, google clearly different, but even more interesting because people are looking for. Who does not want to connect with a former girlfriend, ex boyfriend or girlfriend, etc.. It’s what people wanted most. But I will not stay with this, and as others have stated, we will try to compete with Google and would not surprise me that won. regards,Vudu
I think we’re on the same page, AJ, the one distinction is passive vs. active “searching.” I think Facebook could do a service that lets you actively search your personal and social graph.
I agree. In fact, I’m not sure why they haven’t done this yet.
AJ, I agree with you “In many ways I think Facebook is already a search engine, it’s just not the type of search we’ve grown accustomed to over the last decade.”
-We are not accustomed to social stratification in search results.
The fb open graph currently allows apps, and webpages with their own url to be integrated into fb. Once imported, there is open graph metadata intended to help optimize apps. Each media element; photo, video, pdf etc. also requires a individual url to be included in the graph. Some issues also arise with redirects and child urls. This process is fragmented overall.
Apps built in fb are in an different class altogether, with seamless sharing, optimization for new features like timeline integration. They rise to the top of feeds.
John, it would be great if this were a homogenous solution, it appears URL based media will be weighted to the bottom of this stack, along with hyperlinks to the web.
Anil Dash has a post with added depth and discussion:
http://dashes.com/anil/2011/11/facebook-is-gaslighting-the-web.html
Ah yes, I read that post. I think this has to be addressed
Isn’t the problem with personalization search with Facebook (and other social networks) that the results of a search are based on data from your friends only which one on hand can produce very personal results but on the other reduces diversity?
I think the way to fix that is instrumentation. I don’t think you can do it just through one sided algos.
I know you are planning to write a piece on search+ that includes instrumentation but briefly what do you mean? Also, what do you mean by one-sided algorithms?
Instrumentation means we as consumers instrument our relationship to data, far more than we do now. algos that try to do it for us are one side, ie, they don’t have our instrumentation/
I am not sure an universal search across APP, social and internet is the future direction. Seems to me search and social networks are fundamentally different, so mixing the two may not bring a lot of benefit. When I search, I mostly look for things I am not familiar with, and not the curated information from the past social pages. I am yet to see a benefit from facebook integration with bing.
The ultimate interface for Facebook is to have a link from any item posted to pull up content from your open graph of similar content.
So you see a review of a holiday destination, and one click away are all your friends holiday snaps from that place, their reviews, any travel guides they liked etc.
Plus all relevant personal data from your own apps.
John, you have written/speculated about Facebook getting into the search business this in the past:
http://battellemedia.com/archives/2010/06/will_google_compete_with_facebook_erit_already_is_folks.php
and
http://battellemedia.com/archives/2010/04/its_similar_to_like.php
In both cases, (in the comments) I proposed that Google’s current search business was Facebook’s for the taking, precisely because Facebook is sitting on top of the most important data that matters to most people when they are considering a purchase. Recommendations from friends.
In search, relevance wins. And relevance, by it’s nature, is subjective. It’s personal.
Google will continue to process the lions share of long-tail searches like “What’s the capital of Nicaragua,” but when it comes to considered purchases, where the amount of time between when a person has a question (“Which Atlantic City hotel should I stay at?”) and when they need the answer (in some cases, a few days), Facebook will be able to provide a more relevant experience based on the fact that they are aggregating more personal/social data than anyone else, and also because it’s easy for a Facebook member to “broadcast” their search queries to their social graph, “Which Atlantic City hotel should I stay at?”
This was a revelation that my team pushed hard on back in 2006 and 2007, when I was the head of product development for AOL Search. Personal relevance by way of social context was the guiding principal behind the “Buddy-powered search engine” that we developed (and got patent protection for) back then. The “Buddy-powered search engine” had three main components… First, it searched your social graph to see if your friends had already posted anything related to your query. Second, it optionally notified your social graph that you had just searched for something. Third, it optionally notified you in the future if anyone from your social graph posted anything related to your query.
Personalized results appeared at the top of the page, and Google web search was used as “backfill” down below.. or if there were no personalized results from your social graph, the ten blue links from Google would simply move up to the top of the page.
The “Buddy-powered search engine” was never released to the general public, partly due to short-sighted leadership and partly due to the fact that AOL did not have easy access to content that had been posted to the web by the people on each user’s Buddy List (our plan was to index Flickr, Delicious, YouTube and any other “public data stores” that we could associate back to an AOL username). Facebook has the data, and it doesn’t appear to have the short-sighted leader.
If they do a partnership with Bing for the “backfill” as described above, there really would be no reason for a user to go anywhere but Facebook to search for things, even long tail, non-commercial things like, “how do I tie a fishing knot?”
thanks for reminding me. I swear, without you guys, I’d lose my mind.
John, you have written/speculated about Facebook getting into the search business this in the past:
http://battellemedia.com/archives/2010/06/will_google_compete_with_facebook_erit_already_is_folks.php
and
http://battellemedia.com/archives/2010/04/its_similar_to_like.php
In both cases, (in the comments) I proposed that Google’s current search business was Facebook’s for the taking, precisely because Facebook is sitting on top of the most important data that matters to most people when they are considering a purchase. Recommendations from friends.
In search, relevance wins. And relevance, by it’s nature, is subjective. It’s personal.
Google will continue to process the lions share of long-tail searches like “What’s the capital of Nicaragua,” but when it comes to considered purchases, where the amount of time between when a person has a question (“Which Atlantic City hotel should I stay at?”) and when they need the answer (in some cases, a few days), Facebook will be able to provide a more relevant experience based on the fact that they are aggregating more personal/social data than anyone else, and also because it’s easy for a Facebook member to “broadcast” their search queries to their social graph, “Which Atlantic City hotel should I stay at?”
This was a revelation that my team pushed hard on back in 2006 and 2007, when I was the head of product development for AOL Search. Personal relevance by way of social context was the guiding principal behind the “Buddy-powered search engine” that we developed (and got patent protection for) back then. The “Buddy-powered search engine” had three main components… First, it searched your social graph to see if your friends had already posted anything related to your query. Second, it optionally notified your social graph that you had just searched for something. Third, it optionally notified you in the future if anyone from your social graph posted anything related to your query.
Personalized results appeared at the top of the page, and Google web search was used as “backfill” down below.. or if there were no personalized results from your social graph, the ten blue links from Google would simply move up to the top of the page.
The “Buddy-powered search engine” was never released to the general public, partly due to short-sighted leadership and partly due to the fact that AOL did not have easy access to content that had been posted to the web by the people on each user’s Buddy List (our plan was to index Flickr, Delicious, YouTube and any other “public data stores” that we could associate back to an AOL username). Facebook has the data, and it doesn’t appear to have the short-sighted leader.
If they do a partnership with Bing for the “backfill” as described above, there really would be no reason for a user to go anywhere but Facebook to search for things, even long tail, non-commercial things like, “how do I tie a fishing knot?”
John, you have written/speculated about Facebook getting into the search business this in the past:
http://battellemedia.com/archives/2010/06/will_google_compete_with_facebook_erit_already_is_folks.php
and
http://battellemedia.com/archives/2010/04/its_similar_to_like.php
In both cases, (in the comments) I proposed that Google’s current search business was Facebook’s for the taking, precisely because Facebook is sitting on top of the most important data that matters to most people when they are considering a purchase. Recommendations from friends.
In search, relevance wins. And relevance, by it’s nature, is subjective. It’s personal.
Google will continue to process the lions share of long-tail searches like “What’s the capital of Nicaragua,” but when it comes to considered purchases, where the amount of time between when a person has a question (“Which Atlantic City hotel should I stay at?”) and when they need the answer (in some cases, a few days), Facebook will be able to provide a more relevant experience based on the fact that they are aggregating more personal/social data than anyone else, and also because it’s easy for a Facebook member to “broadcast” their search queries to their social graph, “Which Atlantic City hotel should I stay at?”
This was a revelation that my team pushed hard on back in 2006 and 2007, when I was the head of product development for AOL Search. Personal relevance by way of social context was the guiding principal behind the “Buddy-powered search engine” that we developed (and got patent protection for) back then. The “Buddy-powered search engine” had three main components… First, it searched your social graph to see if your friends had already posted anything related to your query. Second, it optionally notified your social graph that you had just searched for something. Third, it optionally notified you in the future if anyone from your social graph posted anything related to your query.
Personalized results appeared at the top of the page, and Google web search was used as “backfill” down below.. or if there were no personalized results from your social graph, the ten blue links from Google would simply move up to the top of the page.
The “Buddy-powered search engine” was never released to the general public, partly due to short-sighted leadership and partly due to the fact that AOL did not have easy access to content that had been posted to the web by the people on each user’s Buddy List (our plan was to index Flickr, Delicious, YouTube and any other “public data stores” that we could associate back to an AOL username). Facebook has the data, and it doesn’t appear to have the short-sighted leader.
If they do a partnership with Bing for the “backfill” as described above, there really would be no reason for a user to go anywhere but Facebook to search for things, even long tail, non-commercial things like, “how do I tie a fishing knot?”
Makes a ton of sense.
I hope so, Google is now open to sell top rank openly: “Most portals show their own content above content elsewhere on
the web. We feel that’s a conflict of interest, analogous to taking
money for search results. Their search engine doesn’t necessarily provide the best results; it provides the portal’s results. ”
Nothing they do deserves any sane person’s trust, and I mean gmail, analytics, adwords or anything. “Trust me.” No way in hell you unethical pricks.
I want to give Google a week or so to listen to all this and respond. I can’t imagine it will stand mute.
Rumors in the media interwebz and watercoolers is that Google might try to bulldoze through this with a major media company campaign with the “we’re nice and do nice things so trust us.” I doubt it will fly with the influential media but maybe they want to go directly to the people.
I cannot imagine myself working as a Google engineer, everything an engineer stands for has been turned upside down. Where’s the pride of working at Google now when they admit that what they did is akin to selling top ranks openly? (They do “sell” rank but in a more complicated way and not openly–advertisers are on top ten for most lucrative keywords, try a few words and see. They call it “trust” which in reality translates to huge brand or “most likely an advertiser on Adwords.” It really started with the Vince Update and gotten worse as Google’s profits skyrocket.)
Now everyone will look under Google’s hood and ask questions about ads, ad placement, fairness and so on. Every update they will be laughed when they say we “improved the search engine” knowing they rigged it even more to make more cash. Until now they had some sort of way to deny it, now it’s clear.
And it’s hard to say “Yeah I took steroids last year but not this year” especially when you run faster this year. T.A.I.N.T.E.D. I’m shocked no Google engineer has resigned or gone public.
You seem to understand the SEO world pretty well, I’d be curious to hear what you think that world makes of this – from what I’ve briefly read, it seems all SEO folks care about is how to game Google+, now that it matters in results….
Actually, I would have to think that the Facebook search engine could look a lot like this: http://www.rockmelt.com It’s a new browswer built on the Chromium engine with a Facebook platform. It’s all about the socnet way of the web. Enjoyed your article:)
Ah yes, Rockmelt. I’ll grok it again.
Some companies can’t settle a niche/area, they are always hungry for more. Of course this is a good thing, but I’m afraid that there will come a time when we won’t know how to define FB or Google, it’s a search engine, a social site? More is always better, but much more becomes hard to control.
nice post…
great post!
http://dlobi.com
You are on point here. Some people talk about the splinternet. And as the comment by tainted shows and I agree Google is gaming the system. Their last move into social with G+ is good. G+ is great. Not so great is having social results end up high in searches. Yes facebook is big and apps are big. But Google make 37 billion revenue on search. And as long as they stay true to that they can make money. Social is largely hype, facebook and app do generate large amounts of pageviews , but both don’t generate the same profits.
Well, not yet. They are on track to be at about 4-5 billion, if I recall, for 2011. That’s pretty amazing for a five year old company
Don’t try to see god, but work in such a way that God will went to see you.
Sreel Bhakti Siddhant Saraswate Goswami Thakur
“PRABHUPADA”
“Radha Subhash Das ” Bhopal Gour RadhaKrishna Mandir
“Gyan se Bhawan ke bare m pata chal ta hi aur Bhakti se Bhawan ko prapt kiya ja sakta hi .
Atve Gyan se Bhakti sharal hi ”
“Radha Subhash Das” Bhopal Gour RadhaKrishn Mandir.
There’s a Facebook search engine called http://www.FindFacebooks.com that works pretty well. You can search by name, location, interests and other keywords.
Great post John, esp. the part about app/web developers integrating with open graph and then Facebook leveraging the data they receive to personalize the users experience even more.
Thanks Tim!
Now that we know about Graph search, this article was not too far from the point. Man, you saw it coming, though not the full picture. Kudos!!!
Definitely sure it coming!
Holy cow I forgot I wrote that, and one year ago to the day (today!) Thanks.
it is cool to re-discover this article by @johnbattelle:disqus kudos to him indeed!!