Wetpaint CEO Ben Elowitz on the Future of Digital Media
We’ve seen it at Wetpaint, but it’s not just happening for us. Social success and search success now go hand in hand for all web publishers.
If you look at the top 50 publishers on the web, there’s a strong correlation between Facebook traffic growth and Google traffic growth:
For every 1% growth or decline in Facebook visits, the top 50 web publishers in our Media Industry Social Leaderboard saw a corresponding 0.5% change in Google traffic.
Correlation but not causation, you say? I have it on good authority (aka Search Engine Land’s Danny Sullivan) that social signals will soon be the leading factor (if they’re not already) in search engine rankings.
It’s time to drop the notion that an investment in social has to come at the expense of an investment in search. It’s now abundantly clear: social traffic and search traffic go together.
Update: The original version of this post included an incorrect chart. The correct chart is now shown.
This article was published as a guest post at TechCrunch, and is republished here for Digital Quarters readers.
Bing and Google each recently unveiled its own new search interface, designed to better intuit your intent and help you get to the one best answer more efficiently. And they’ve made it ever more clear that search is heading straight for a merger with social.
The changes are smart. Google’s knowledge graph is useful – when I search for certain things, I just want a cheat sheet. What is Faraday’s Law, again? What exactly is a geoduck?
But Bing’s new feature – “people who might know” – is even smarter. This is the first major attempt at a merger of search and social – unless you count Search Plus Your World, which I don’t – and this is undeniably the way we’re headed. There’s a lot of information on the internet, but getting the right info from the right person is still a huge, and mostly unsolved, undertaking. Nobody knows the answers better than, well, somebody who knows the answers. And so much the better if it’s someone I trust. (Thank you, Jeff, for the Singapore recommendations!) The fundamental insight is that when I ask a question, there are lots of ways to help me find the best answer. If you don’t have it, point me in the direction of someone who does. Don’t make me ask the same question in a million permutations and sift through a list of 20 possible right answers every time.
What’s more interesting is that this is the biggest step forward we’ve seen since search results started looking 12 years ago the way they still do today (just with more images and toolbars now – exactly what Google got rid of back then!).
Stagnation followed by the springtime of innovation is probably the surest sign that a major disruption is imminent. (And if that weren’t enough, just think of how much Facebook’s stock price would rise if they captured even a small share in search.)
What’s the endgame? In 10 years, I’ll still need recipes for dinner. And recommendations for hotels in a new vacation spot. And to find something to do on the weekend. I know how I would make these decisions today, but how will I make them in 2022?
The true merger of social and search will look nothing like the search we know today. I don’t even think we’ll call it “search.”
The social search of tomorrow will be more like a combination of a whip-smart personal assistant and an intuitive, considerate significant other. But one who’s exponentially more efficient and who doesn’t mind being woken up at 3am. (I’m lucky, but not THAT lucky!)
Let’s put on our future-goggles and imagine how a fully social, personal-data-powered search would change our day-to-day:
Proactive: It’s Tuesday night and I’m hungry. Luckily, my mobile knows that I just got a CSA box containing sweet potatoes (Full Circle Farm’s Facebook integration), and that I tend to eat at home on Tuesdays (according to my historical pattern of check-ins). It also knows that it’s cold and raining outside. Before I’ve gotten around to opening a cookbook or the Epicurious app, my mobile pushes me a sweet potato soup recipe that my certified-foodie friend raved about on Facebook last week.
Personal: Arrive at the Sao Paulo airport and search on my mobile for the city’s public transit map. My device knows that I’ve never been there (even though I bought a phrase book on Amazon last week), and it also knows (from scanning TripAdvisor comments about Sao Paulo buses) that the public transit is impossible to navigate for newcomers. While the map is loading, a message appears gently encouraging me to consider a rental car instead – there happens to be a great deal on an Audi (my favorite(!) as noted on Facebook) at the rental counter 10 feet away. Talk about targeting!
Social: Florence and the Machine is touring in New York, and I’m dying to go see them. I called the usual suspects, and they’re out of town during the concert. The only thing worse than not going is going alone. But who else do I know who loves them like I do? That’s a lay-up for a socially powered search if ever there was one. Two words: “Jason Hirschhorn”. Is that so hard?
There are a hundred other decisions that would be made immeasurably easier with the help of a really good personal assistant – one who knows your schedule and your preferences (and the schedule and preferences of your friends and family); one who has excellent research skills and can track down the appropriate expert on any issue. (But no, I’m sure it still won’t replace Larisa.)
Most of us don’t have personal assistants. But we have left a heck of a trail of our interests, associates, habits, and dislikes. It will take some algorithm to turn that trail of behavioral and social data – combined with the wisdom of topical experts and the vast repository of information that is the internet – into a set of smart, personalized answers for you and me. But that’s why Google and Facebook and Apple hire engineers with such big brains.
And, surprise!, the better they understand our brains (read: intent, context, and relationships) the better the match they can serve up to an advertiser. And that means an outrageously good search not only retains audience better, but would improve ad rates.
We’re on the verge of shifting from a search model in which the user is still doing all of the heavy lifting to one where powerful algorithms enable our devices to anticipate our needs and do most of the sifting and evaluating for us. In the meantime, though, we’re stuck in a “hairball of complexity” (to borrow Adam Richardson’s TV industry analogy) while the industry struggles to find the way from A to B.
The key is in having software that recognizes us as whole people. (And isn’t that exactly the promise of social?) Now search is undergoing a massive transformation from receiving input in the form of queries – each independent and atomic – to understanding its input in the form of people, who have personal history, context, and relationships. That means delivering the right result depends on who is asking. Which is sooooo true. I don’t like the same music as my teenage niece, and she doesn’t like the same restaurants I do. Why should we both get the same search results?
Apple’s Siri is certainly the closest, at least in spirit, to the eventual reincarnation of search as personal assistant, even as its true capability has far to go. The voice-activated question-and-answer experience is light years ahead of the long list of links on a page that still defines search on Google and Bing. But the trick that remains is to gather, combine and analyze data from myriad sources – social interactions, behavioral data, expert opinions – and deliver it back to the user in a way that makes decision-making more efficient than most of us can imagine.
With all of that time I used to spend inefficiently making decisions suddenly freed up, what will I do? I’ve been meaning to plan a trip to Sao Paulo….
This article was published as a guest post at XConomy, and is republished here for Digital Quarters readers.
I’ve been taking in Google’s recent release of “Search, plus your world” (or SPYW as the cool kids say) over the last several days, reflecting on what it means for Wetpaint and other media companies; but perhaps even more importantly, deeply understanding what it indicates about Facebook and Google themselves. As we all know by now, these most recent changes are meant to make its search more personal by up-weighting social activity in its algorithm, and using each person’s own position within their circles to determine relevance.
You might think that I would be one of the first to jump in the game with Google. After all, my company Wetpaint has been making a massive investment in distributing our content via other social channels, particularly Facebook. We’ve been seeing massive returns. And, I’ve even gone on a limb to predict that Facebook should be implementing its own Web-wide search this year.
Still, when it comes to playing Google’s social games, so far I’ve advocated staying on the sidelines of all their social venues—even their recent business pages. That’s been because even though the stadium lights are on, no one is on the field. More specifically, even though Google has 90 million registered users of the service, we see very little activity of significance among our target audience. But with its new SPYW changes, the question is: Has Google indeed forced companies’ hands?
Unfortunately, they have. And, in doing so, it marks a milestone in the changing mentality of Google. The search company’s great innovation—using the signals of the Web to best determine what the audience really wanted—has now been subverted. The company’s originally unshakable-seeming ethos of mechanistic neutrality has slowly, slowly, slowly, and now all of a sudden given way, and the new precedent is to favor its own business interests over those of the audience.
The result, like it or not, is that companies that rely on search for traffic must hear and obey loud and clear Google’s message that Google will favor those that favor it. It’s a dirty truth, and one far more chilling than the other more technical biases of its algorithm before.
Google has already started infusing search with the content that’s been blessed via Google+. Do a search for “New York Times” and you’ll probably find the New York Times plus.google.com page as the second search result. Search for “Mark Zuc” and you’ll likely see Zuckerberg’s Google+ page (despite the irony) populate as an option in the Google Instant choices.
I haven’t seen this bleed over to news stories yet, but I believe that it’s coming. Soon you’ll do a search for the latest headlines and your search results will be chock full with musings from your friends and non-friends inside Google+.
Google+ may not take off as a real social network, but Google has indicated that it’s throwing its full weight behind it anyway to make the best of what it’s got. Even if consumers don’t adopt it en masse, whatever activity is present will pepper the famous algorithm’s search results.
The irony here is that Google’s pivot toward a social search belies how important that social data is. The company is putting its lock on search at risk to gain a chance at a foothold on social. But what really comes through to me is that a great social search can be a winning product—if it’s populated with the right social data. So far, Google’s is not.
The question is—if that’s what I’m after—won’t I still just go to Facebook, where all my friends actually are (and which Google has adamantly cut out of SPYW)?
While SPYW does force publishers to support Google’s social network, fortunately it will be a temporary sacrifice from publishers during this period of transition from these days of search to a socially wired world. And that forthcoming world looks increasingly like it will be wired not by Google, but by its arch-enemy Facebook. Indeed, by corrupting the quality of their search product, Google may have just opened up a clear product entry into search for their rival as well.
Next year, search advertising will be a $15 billion market in the U.S. alone, growing by 14 percent, according to eMarketer. And, if Facebook can capture half the share of that market that Google has today, it could easily add an extra $25 billion or even far more to its value.
For most any CEO who could have even a modest chance of succeeding at it, that payoff would be reason enough to take a serious look at entering the search category. And yet, while I’m sure he wouldn’t scoff at the extra revenues, profits, or valuation, I suspect that Mark Zuckerberg finds something else far more motivating than just increasing the financial value of his company.
And that’s what will propel him next year to make a completely disruptive entry into the search category.
So if it’s not for the financial value, then why am I so certain that Mark Z. will make a play for Google’s home turf?
It’s because it’s so irresistibly good for his users. And that’s the most important principle that seems to guide his product development.
With that in mind, here are the five specific reasons why Facebook should enter search next year:
1) To make Facebook the ultimate home page. Consumers make Facebook their home base. Half log in every day; and users come to Facebook 70 percent more times per day than even to Google. They stay twice as long as even users of Yahoo’s vast network of email, content, and more. Facebook has become the Connected Web’s de facto operating system. But right now, its “start button” is limited to what other people put in your newsfeed. Part of being home base is being a launching pad to go anywhere you want. So Zuckerberg will need to give users a great connection to the rest of the Web – whatever their intent.
2) To fix a broken feature. Facebook has a search feature today (powered by Bing); and a few people already use it for Web-wide search, even though it isn’t very good. It needs significant upgrading, and Zuckerberg knows it. Having a feature this important be this incomplete creates an unacceptable user experience. It must be fixed.
3) To improve people’s life online. Facebook has an enormous data set that it can use to deliver better search results than anyone on the planet. Facebook can see everything that Google can see in terms of pages and links, but with a whole extra dimension of human connection that is impenetrable to Google. Facebook knows what your friends like, and what people like you like. And it knows the difference between real interest and spam. Translating that knowledge into great results will improve online life for his users.
4) To fully connect the world. More than anything, Zuckerberg and his company’s DNA are all about providing services to connect users to each other and, increasingly, to the world at large. Serendipity and sharing aren’t enough: sometimes people know what they want to find. Facebook must have a search feature to fully enable connection.
5) To add to his immense data set. Search will not only help users; its users will help Facebook. Specifically, it will provide Facebook with even more data about what people want so Facebook can further personalize itself for everyone. Go ahead and cue the creepy privacy music, but remember that so far most users have been happy to make a privacy tradeoff to get valuable personalized service.
With Facebook Connect, Open Graph, and Like buttons, Facebook has already shown its vision to fully connect to the rest of the Web. The next step is to help people better access it.
Facebook began as a social application, but it’s now in the process of becoming a Social Operating System for the Web at large. Offering world-class search is the next step in its evolution as that “Social OS.” The Web is now organized around connected people, not documents – and Facebook is the OS that links those people together.
Once fully connected, can you imagine how Zuckerberg must think about a Web all wired-in through Facebook’s central hub? He’d know the time spent on every page; the usefulness of every link; the patterns of every user. He’d have a real-time system that provides feedback on every recommendation. You know what’s cooler than a billion connections on the Web? How about a quadrillion!
The value of that data will be immense in making recommendations to users, serving advertisers, refining search itself, and enabling next-generation social applications. It will give Facebook a competitive advantage over every other Internet company in building a map of where the gold is buried – in the form of the content each individual user wants – among the trillions of pages on the Web. But more importantly, it will allow Zuckerberg to serve his users.
The idea of a socially-powered search is not brand new for Facebook. Bing and Blekko have both incorporated features that bring your friends’ Facebook content into the search results. And while that is one modest way to improve the search, its impact pales in comparison to the full potential of what Facebook can do to help you by fully exploiting its social data set: It can individualize search results just for you, by using not only data about you and your friends, but by using the full dataset of people you haven’t even met yet.
Let’s look at it competitively. Google and Bing have, with limited exceptions, held themselves to the standard that the results should be the same for everyone because they work in an anonymous environment. A friend from Microsoft tells me that Bing has a rule that, with the exception of bucket tests, the top ranked result must be the same for everyone. This rule, he says, was copied from Google – where it fits well with Google’s increasing positioning of itself as the great defender of identity control, compared to Facebook’s ethos where everything is public. But that differentiation hands Facebook an incredible opportunity: in the Facebook environment, it’s not only accepted but expected that everything you do is customized for you alone.
Can you imagine the power of combining Amazon-like personalization with Facebook’s deep dataset to offer better results?
That’s why beyond just improving a search algorithm, Facebook’s greater opportunity is in redefining the category. The last decade of Web use has been defined by Google’s clean white splash page with a single query field, and the 10 blue links which follow. But just as that approach from Google displaced the prior generation’s directory pages, it’s time for a breakthrough experience. And Facebook is the natural player to provide it.
I’m sure the engineers at Facebook are already visualizing what search could be in a fully connected world. Searches could be proactive, prompted by items shared by friends, rather than awaiting a text field completion. Searches can favor brands and publications that you like, or your friends like. But most importantly, searches can be predicted based on people like you, people who are located where you are, or people with similar interests, profiles, and behaviors, without you ever even knowing them. All of these are ways that Facebook can fundamentally redefine search, thanks to its knowledge of each user’s identity, interests, and behaviors.
But building a search engine that takes (a difference-making) advantage of the social graph takes lots of time and money, as does building a new operational infrastructure, Web crawler, and advertising engine to support it. And, even more significantly, this is one where Zuckerberg will need to get the privacy implications right from the start. Facebook is currently building its rep with major advertisers on its social network – and that’s a great start, because that will provide a captive customer base to transition into its search engine right at launch.
A competitive search engine is one of the most ambitious projects you can imagine – the degree of difficulty is mind-boggling, and the cost is hundreds of millions or more. For Facebook to best Google, it would need to catch up in substantial ways before it could shoot ahead of the leader, even with its valuable dataset. But that’s only an impossible challenge if it has to do it all alone.
And Facebook doesn’t have to.
It already has an alliance with the #2 player in search, Microsoft. And – in the way of “the enemy of my enemy is my friend” – it has a common interest in outperforming Google. And Facebook and Microsoft have enough separation between their businesses that they could complement each other rather than compete. Indeed, Facebook’s increasing strength in its advertising engine could be a huge lift to Bing’s struggling monetization – offering hope of raising Bing’s monetization toward Google’s levels. The two truly are more valuable together, and it’s no surprise that smart people have begun to speculate on a Bing-Facebook combination, a step beyond a partnership. Working with Bing for its search entry could save Facebook billions of dollars of initial R&D and speed its entry into the category by years – and by many dozens of engineers. And any agreement they’d sign would likely still give Facebook the option to create its own search engine down the road.
Regardless of how Facebook structures its efforts – and with whatever degree of help it gets from Microsoft – it will be able to create a search capability that will be significantly different from anything we’ve ever seen. And it will shake the tectonic plates underneath Google’s Mountain View headquarters, even as it vies to earn users’ adoption with better, more personalized results.
Google will not perish in the digital earthquake without a fight, though. Its recent Google+ launch, for example, shows just how boldly Google intends to enter Facebook’s home territory. That, of course, makes it even more imperative for Facebook to counter-invade by pushing into search.
Looking forward, it’s clear that search and social won’t always occupy separate spaces. Indeed, for consumers, over time, they will converge; and the blended (or, just as likely, reimagined) product that emerges will serve as a home base that will serve as a jumping-off point to everything that’s important and relevant on the 21st century Web.
It’s fascinating, and it’s all about to unfold. In the meantime, while Zuckerberg quietly forges ahead, and readies Facebook’s game-changing search entry, Eric Schmidt, Google’s former CEO, is publicly lamenting lost opportunities to catch Facebook. The diverging fortunes of these two digitally defining companies could not be more apparent right now.