Cost-per-action social media and search advertising service Taykey has just closed a $9 million second round of funding led by Sequoia Capital.

The company tracks social media mention and search trends in real-time to take advantage of urgent advertising opportunities and interest-based targeting parameters to power Facebook, Twitter, and search campaigns. Here’s a closer look at what it provides to advertisers on and off Facebook.

How Taykey Works

Let’s take a look at an example of an advertising campaign powered by Taykey. A singing contest television show such as X-Factor could hire Taykey for a month-long campaign on what amounts to cost per Facebook Like model, where the client pays a price per Like and sets a goal of the total Likes they want. Taykey would then monitor social networks, news, blogs, and searches for trends in the behavior and interests of the show’s target age and location demographic. It could determine what other TV shows or musical artists the audience Likes, and then run a series Facebook ad campaigns for the X-Factor Page targeted at people with those Likes.

Taykey co-founder and CEO Amit Avner tells us that “if X-Factor judge Paula Abdul falls off the stage, we’ll know in five seconds and go buy ‘Paula Abdul’ Google search keyword ads” to preempt the oncoming rush of searches for that keyword. Taykey might also purchase Twitter trending topics, or ads on Bing, Myspace, Digg in an effort to drive its own cost per Facebook fan as low as possible to make the maximum margin on the deal with X-Factor. Otherwise, X-Factor might just target 18-35 year old women, whereas Taykey would target those with interests related to the show, such as those who Like competing show American Idol.

Taykey says its patent-pending algorithm mines data from across the web, deduces keywords and sentiment, and maps the data to demographic and psychographic profiles. It then specifically targets those with the right profile, relieving brands from having to constantly discover new targets. Without using cookies or tracking of individuals, it shifts spend from one trend to the next attain the optimal CPA.

Avner explains to us that brands advertising on Facebook often target an age, gender, and location demographic that is too wide and unfocused, leading to lower conversion and rapid burnout. Taykey pinpoints the interests of these audiences to run a series of campaigns that keep conversion rates higher over time than more general campaigns. For instance, instead of targeting 18-24 year old males in New York City that Like ‘hip hop’, Taykey would determine specific artists such as “Jay-Z”, or television shows such as ’106th & Park’ to target the fans of.

Taykey and Real-Time Facebook Advertising

The 19-employee Tel Aviv-based company was founded by three former members of the Israeli Defense Ministry’s intelligence arm, and has now secured a total of $12 million in funding. Several Fortune 100 companies have already run campaigns targeted by Taykey’s algorithm, including Pepsi, for which it attained 46,000 Likes in two days at half the projected spend.

It will use use the new round of funding, joined by Softbank Capital and Crescent Point, to hire engineers in Tel Aviv, Israel and to build out a New York sales office.

The “Related Adverts” real-time advertising system Facebook is testing that displays ads related to the content of a user’s most recent status update or wall post could be very useful to Taykey’s business. “We’d love to get in on it as soon as possible” says Avner of the beta product that doesn’t allow advertisers to choose if or which traditional interest-targeted Facebook ads are displayed in real-time.

As more brands realize the concrete value of Facebook fans, CPA ad services such as Taykey will become crucial to attaining large volumes of fans at the lowest possible price. While more well established Ads API tools and services will likely continue to manage much of the Facebook spend of the world’s biggest brands, real-time focused advertising services can complement a marketing mix by exploiting fleeting low-cost pockets of conversions.

Earlier this week, Microsoft’s search engine Bing significantly deepened its Instant Personalization integration with Facebook. It launched several new social features for its web search, shopping, travel, and toolbar products; and it began triggering the display of social content much more frequently.

By improving result quality with both the data of a user’s friends and the Facebook user base at large, Bing is positioning itself as the search engine of choice for when people need social reassurance to make a decision.

We sat down with the Director of Bing, Stefan Weitz to ask him a few questions about the motives behind the update. We’ll follow that with an in-depth look at the product changes.

Q&A on Social as the Future of Ecommerce Search

Inside Facebook: Why is Bing pushing to get social integrations so deep into decision-based search?

Stefan Weitz, Director of Bing: Decision search is moving from exploration to active conversation because 80% of people delay making a purchase online until they can talk to a friend.

Core search stuff has been taken care of by intelligent organization. But how do people sort through all the links and make a decision? You do all the research, but at the last minute you walk away from the purchase process because you’re not convinced until you get a social recommendation.

When you want to tap into that info you have to go to a bunch of different places. Whether it’s just Facebook that solves that problem, or if it’s Quora and other sites, behavior is already moving in that direction (of seeking advice online rather than offline). But no one does a good job of pulling it all together into search.

IF: What has changed that’s made this possible but also necessary?

SW: Stuff that was previosuly in your brain is now in a format that machines can read. Friend connections are a new way of thinking about ranking search.

Meanwhile, humans are creating 5 billion gigabytes of data every two days, and machines are losing their ability to categorize it all. How can PageRank handle a Yfrog image? It probably doesn’t have a title, or caption, or anything else that could help index it. But if a friend Likes it, that’s important.

IF: Why is the social content appearing more frequently now?

SW: Honestly, it was light before. You didn’t see much of it.  Now it’s gone up a ton, you’ll see it a lot more. It triggers more because we have higher coverage [across products]. Its more than just Likes now. We think of people as having characteristics and attributes, not just actions. Now we’re considering what other meta data can we use that people will give us access to so we can continue to personalize search.

Improvements to Bing’s Existing Facebook Integration

Bing began its Facebook integration by indexing Page updates and publicly visible links posted by users in June 2010. In October of that year, it partnered with Facebook to offer Instant Personalization of Bing search results so users could could see Likes by friends of search result objects, and their network connections to Facebook users found through a name search.

However, the search result Likes were displayed very infrequently, so some hardly noticed the change. Microsoft also released the Bing Bar toolbar, which allowed users to view the Facebook news feed and their notification from any tab, but it didn’t offer an easy way to share web pages and links copied into its Facebook publisher weren’t formatted as they are on Facebook.com.

Bing has now fixed these issues and greatly expanded the functionality of its Facebook integration. As Weitz said, social content now appears in search results much more frequently. Meanwhile, the Bing Bar now has a “Universal Like Button” — a one click way to share the currently viewed webpage to the news feed with the same rich story formatting as if one had pasted the link into the Facebook.com publisher.

Search With More Social Content

Along with more frequently showing you Likes by friends of things represented in top search results, Bing now actually personalizes the rank of results based on these Likes. Weitz tells us that “based on the actions of friends, results that would be on page two or three are pulled onto page one.” Even if no friends have Liked a result, in some cases it will display the total number of Facebook users that have Liked it, helping users make decisions about topics outside the expertise of their network.

In one of the most useful new additions to Bing, frequently Liked webpages from within popular websites will appear beside their Like count underneath a result, allowing users to sift through today’s content heavy blogs and community sites. For instance, articles from a news site or recipes from a cooking site that have been Liked by friends or many other users are now much easier to discover.

Related updates from Facebook Pages now appear in general search instead of being isolated in the dedicated “Social” search tab. As Pages become a sort of news ticker and deals distribution platform for many brands, Page updates are becoming valuable content to surface.

Bing Augments People,Travel, and Shopping Search With Friend Profile Characteristics and Sharing Options

While before users could search for people through Bing, the results were no more helpful in finding the right “Bob Smith” than Facebook’s own search engine. Now people search results display profile data visible to the searcher, such as current city, workplace, and education history if its public or is to visible to “friends of friends” and they have a friend in common.

Similarly, Bing Travel searches will display the friends that live or previously lived in a city that’s been searched for. In an innovative use of the ability for owners of Like buttons to publish news feed updates to those that click them, users who Like flight results between two cities will receive feed stories about about deals on those flights. Users can also share Travel Wish Lists with friends. Bing Shopping has also rolled out it’s previously announced shareable product comparisons so users can seek advice from friends.

Where Social Search Can Go From Here

Bing’s latest social additions could make it a sensible tool for people inquiring about things outside their comfort zone. If I don’t know anything about LCD TVs, but want to buy one, Bing could help me discover reputable brands and popular sources of tech hardware reviews, or share the choices I’m comparing with my friends.

There’s still more to do, though. For instance, I might see that one friend Likes one tech hardware review blog, and another friend Likes a different blog. Perhaps Bing or another social search engine could analyze the full set of Likes of those friends, and determine if one is an expert on technology because they’ve Liked several other tech brands or publications. This friend’s Like could then be weighted more heavily or shown more prominently.

There’s also the question of those who really don’t want a social search experience. Now, even if a Bing user doesn’t have a Facebook account they’ll still see anonymized data about the Likes of the general Facebook user base. By providing an option to completely opt out of seeing Facebook data in search results, Bing could continue aggressively integrating social data without alienating those who don’t believe in the wisdom of the crowd.

Tagging was arguably the feature that made Facebook the biggest photo site in the world and seeded the idea for creating the platform.

Now the company has finally won a patent for it.

Nearly five years after the company originally filed for the invention, the U.S. Patent and Trademark Office gave Facebook a patent protecting the ability to select a region in a piece of media (like a photo or video) and associate people or other entities with it. Mark Zuckerberg, longtime designer-turned-product architect Aaron Sittig and former Facebook engineer Scott Marlette were credited as inventors.

Zuckerberg has long talked about photo tagging as the innovation that helped him and other early Facebook employees initially conceive of the idea for the platform. The company did a competitive analysis of all other photo products out on the web and while Facebook didn’t offer features like high resolution or printing, it still outcompeted rivals simply because it centered its product around people, and not around technical capabilities. Last year the company said it was seeing more than 100 million photo uploads a day. It has not updated that statistic since.

Because of photo tagging’s initial success, the company started thinking about other products and verticals that could be reinvented around social behavior. When it became clear that with limited engineering resources, the company wouldn’t be able to create every single possible idea on top of the social graph, it opened up the ability for other third-party developers to do so.

The company won one other patent in the last month too: the ability to give gifts in a social networking environment. This one was credited to Jared Morgenstern, who is a product manager on the games team. Gifts were ultimately deprecated in 2010, but virtual goods have become an indirect source of revenue for the company through its currency Credits.

Facebook also applied for four search-related patents in the last month that control how results are shown to users based on their social proximity to the information or how often they access it. All of those patents are credited to Christopher Lunt, Nicholas Galbreath and Jeffrey Winner.

Facebook now splits the instant typeahead search results that appear in the drop-down beneath its ever-present search bar into profiles, Pages, apps, groups, Events, and Questions. This helps users find the most relevant results, regardless of category.

Category-sorted results could also help users discover new content they weren’t looking for but that relates to something they were, like a soccer game when they were searching for a soccer Page. Finally, the change has SEO implications, as appearing in the first few results within a specific tip is more important now, and Pages and apps can score incidental traffic by having titles similar to common names.

The different result categories are ordered according to the relevance of the matches, meaning users could see any category first. Since the categories aren’t weighted according to their popularity, the ordering can be a bit awkward at times. Users are more likely to be searching for profiles, Pages or apps than groups, Events, or Questions, yet the first result may still be a Question.

Though this unweighted ordering isn’t necessarily best for the user experience, it may expose people to some of Facebook’s lesser-known core apps.

Its now even more crucial for applications and Pages to attain one of the top spots for a given keyword. It also incentivizes them to name themselves after or something similar to popular names. For instance, an app called “Just For Fun” might receive traffic from people searching for someone named “Justin”.

Facebook has been gradually improving its internal search feature. It began surfacing Liked Open Graph-objects such as news articles in August. If it can continue to improve its internal search, users will be more likely to use it instead of web search engines like Google in order to find things on Facebook.

We had a few interesting posts this week pertaining to partner engineering and management and global Credits marketing.

The position for Partner Engineer of Ads in New York is set to work with Facebook’s strategic partners as they integrate the social network into their web sites, applications, devices and more. This person will travel and is expected to have substantive technical knowledge.

The Partner Manager of API Programs will be based in Palo Alto, Calif. and works to “accelerate advertiser growth, satisfaction and value by guiding tool developers to build value-added third-party applications” to support Facebook’s growth. Substantive technical knowledge is also required for this position.

Facebook is also seeking a Global Credits Marketing Manager to lead its Credits marketing to developers, partners and users from the Palo Alto office. Part of the job is brand awareness, including with retailers and work to create new markets for Credits. The ideal candidate would have about a decade of marketing communications experience in technology and a strong understanding of techology.

Facebook also advertised a Product Manager of Search position based in its Palo Alto offices. It’s not clear if this position will be augmenting Facebook’s relatively straightforward search results on its site, or developing something more full-fledged.

Facebook added a handful of account executive and manager positions in its Austin, Palo Alto, Madrid, New York, London and Chicago offices. A few more positions were added in Dublin, Singapore and Hyderabad.

Jobs that were removed from the Page, and thus possibly filled, ranged from an Associate for Corporate Communications to a European Facilities Manager to an Analyst of Platform Operations – Arabic. A handful of jobs in Facebook’s Dublin, Ireland offices disappeared, as well as some account executive positions, a SWE1101M Software Engineer position was filled, as well as a technical recruiting job. Some other jobs no longer appearing on the Page: Product Manager, Search Relevance (Ranking), Manager of Global Compensation, Diversity Programs Manager and finally, Partner Development Manager of Mobile in Hong Kong.

For more Facebook-related jobs, check out the Inside Network Job Board.

Experian Hitwise released statistics yesterday indicating that “Facebook” was the most frequently searched term for the second year in a row, besting YouTube, Craigslist, and MySpace. Among the top ten searched terms, four were variations of Facebook, including “Facebook login” showing that users frequently search rather than entering the site’s URL or clicking a bookmark when they want to browse the site.

The rankings of the top searched musicians closely mirror our PageData charts of the most popular Pages, showing that Like count may be an accurate measurement of public interest. Facebook also overtook Google for the first time to become the most visited website of 2010.

Social networking sites claimed 4.18% of the top 50 searches, demonstrating that the public is very interested in using and learning about ways to interact with friends online. Facebook, and variations like “Facebook.com”, made up 3.48% of U.S. searches among the top 50 terms, a 207% increase over 2009, outpacing Facebook’s monthly active user growth for the year.

The climb of the term “Facebook login” from the ninth to the second most searched term means users are interested in a quick, simple way to access their account. Facebook has been trying to help users achieve this by prompting some to add the site as their home page.

All five of the top searched musicians, including Lady Gaga and Michael Jackson, are also among the top 30 Facebook Pages and top 11 musician Pages on Facebook according to our PageData tracking service. While Like counts can be artificially increased through marketing and Like-gated promotions, the correlation in the statistics shows that Facebook Page size may be an accurate gauge of public interest for some categories.

Google led Facebook in terms of total visits across its owned web properties with 9.85% of all U.S. visits, summed from its search engine’s 7.19%, YouTube’s 2.65%, and other sites including Gmail and Google Maps. Facebook.com alone brought in 8.93% of all U.S. visits, meaning it might one day exceed Google in this category if it expanded its service offerings.

Hitwise stated that “the top 10 Websites accounted for 33 percent of all U.S. visits between January and November 2010, an increase of 12 percent versus 2009.”, showing the biggest sites in several categories are pulling share away from their smaller competitors. This consolidation could be fueled or signaled by the numerous acquisitions by web giants like Facebook and Google in 2010.

Wildfire launched a social media measurement tool called the Social Media Monitor today. The new service is set up on a web site to allow visitors to compare the performance of Facebook and Twitter accounts and will soon also include Places check-ins and overall fan/follower engagement.

In an example of the Social Media Monitor’s utility Wildfire recently ran a study comparing Target and Walmart. The Monitor showed that Target beat out Walmart in terms of captive Facebook audience during the summer, has extended its lead in recent weeks, particularly around Black Friday, putting Target 600,000 fans over Walmart.

Using the Monitor is very simple. A user just goes to the web site, enters the Twitter or Facebook URL for a particular company/brand and voilà, information that includes growth over the past 7 days, month and 3 months appears, along with a graph and the total number of fans/followers.

To help visitors to the Monitor take their social media reach to the next step, Wildfire also provides links to its other services along with these metrics. Learn more about strategies to increase your social media reach with the Inside Facebook Marketing Bible.

[Image Via Wildfire]

Facebook has partnered with Yandex, Russia’s largest search engine, to index updates to Pages and show users of the search engine their Facebook notifications. The integration is not live yet, but will augment Yandex’s existing index of twitter.com, livejournal.com and Russian blog platform liveinternet.ru.

Yandex will show different Facebook content in search results than Microsoft’s Bing, which joined the Instant Personalization program this month. Bing shows Likes on search results, while Yandex will include Page updates in a similar fashion to search engines including Booshaka and Open Facebook Search. Facebook will also display a widget on the Yandex front page, providing constant exposure for the social network.

Facebook recently increased its focus on the Russian market, signing deals with Russian mobile carriers to offer free access to Facebook. Despite having to compete with local social networks such as Vkontake and Odnoklassniki, Facebook is now growing relatively quickly there. According to The Facebook Global Monitor, part of our Inside Facebook Gold membership service, Facebook increased it’s Russian audience by 16.8%, or 228,960 users in September 2010. Still, Facebook has only registered 1.1% of the Russian population, so the Yandex partnership presents a new way of gaining exposure to Russia’s 60 million internet users.

Instead of crawling Facebook to index social content, Yandex will receive a syndication feed of Page updates directly from Facebook. This reduces the lag between content being posted and being indexed. The press release also mentions personal search, but doesn’t explain whether Yandex will show Facebook user profiles or Page profiles as results to searches for names.

Javier Olivan, Facebook’s Head of International, says “this partnership will bring a lot of value for our Russian users”. The partnership will get the Facebook name and content in front of millions of Russians. This may help Facebook to accelerate growth in the area in exchange for giving Yandex faster access to data which is already public.

4Loot allows Facebook users to win coins which can be redeemed for Facebook Credits by using the Yahoo!-powered 4Loot search engine. Valid searches count as sweepstakes entries, and users are randomly selected to win coins. Since many of the search results are sponsored, coins awards are few and infrequent, and Credits cost many coins, 4Loot is an inefficient way to earn Credits compared to other Credit earning services like ShopKick, ifeelgoods, and Game Coins.

Currently, the only method of earning coins is through the 4Loot search engine, and the only prize they can be redeemed for is Facebook Credits, but the company says more earning methods and prizes are coming. The site has been gaining traction, and even temporarily buckled under the traffic, possibly due to a post on the Facebook Credits Page touting the service.

Users are randomly awarded between one and thirty coins if their valid search (common websites names, random character strings, repeat searches, and searches which produce zero results are invalid) is randomly selected by 4Loot’s algorithm. Clicking through results, including sponsored results, does not affect a user’s chance of winning. Other restrictions include one account per household, and only US residents are eligible.

Credits cost between 20 and 32.5 coins per Credit at the current $0.10 per Credit exchange rate. The site frequently cites top Zynga and PopCap games where the Credits can be spent.

Users must log in through Facebook and give 4Loot several permissions. After seeing a prompt explaining how to win and redeem coins, users are presented with the search bar. The first page of search results shows eight sponsored results which are difficult to discern from the twelve organic results. Subsequent results pages show slightly fewer sponsored results.

Once a user has won enough coins, they can be redeemed for different quantities of Facebook Credits on the Loot tab. The leaders tab lets users see who has won the most coins that week, as well as a big photo of the weekly bonus winner of 50 coins. Users can invite friends using the multi-friend selector on the Friends tab, and see their past winnings and purchases on the Me tab.

While the interface is simple and intuitive, few will be interested in switching to 4Loot’s degraded Yahoo! search engine for the low chance of winning one coin, or $.005 worth of Facebook Credits. 4Loot’s reward to investment ratio is far below that offered by other Facebook Credit earning services. Therefore, the only foreseeable users are minors without credit cards or steady incomes who don’t want to visit retail locations, install apps, or provide contact information through other services to earn Credits.

Facebook friend relationships and Likes will shortly begin appearing in Microsoft’s Bing search engine, the companies announced today. In addition to the 1,000 other factors that Microsoft says it uses to determine results, there will now be items on the results Page showing friends and Liked Pages that are related to search queries.

These social additions could improve the value of Bing to users, helping them find information about people or other things — restaurants, clothes and other goods — based on existing Facebook data. See our coverage from earlier today for more details on the product.

The bigger story, that many are observing, is that Microsoft and Facebook are collaborating to compete with Google on its home turf. While that is clearly the case today, the situation for these companies is more complex.

Microsoft invested $240 million in Facebook back in 2007 at a $15 billion preferred stock valuation, and it got a strategic partnership as part of the deal. While the valuation was widely questioned at the time, Facebook’s valuation has now potentially more than doubled (according to limited sell-side data from some secondary markets), and Microsoft is in position to work out product relationships like the Bing one today. It has already been pushing out Facebook-related updates, actually, such as when it began surfacing Facebook public status updates in June.

On the other hand, Google has had limited access to Facebook data, most notably introducing public Page information to results back in February. Although user names, photos, friend relationships and Likes are all public information, Facebook controls how third-parties, including search engines, collect it for their own purposes. It blocks most third parties from scraping the site (at least as best it can) while it works out deals with partners like Microsoft. So Google has relied on its own social data, as well as more from Twitter and other partners, to help personalize search results.

If the additions today — or any future ones — significantly improve search, then Google will feel even more pressure to try to collect its own social data, or cut a deal to get Facebook’s.

Not Just About Search

But what are Facebook’s long-term plans for its data and search engines? As chief executive Mark Zuckerberg said today at the launch event, “We’re trying to build a platform, so fundamentally, this is not about working with a single company. Over the long term, we would love to work with everyone.”

To rephrase that line, whatever Facebook’s relationship with Microsoft, and whatever interest it has in disrupting Google, what it really wants to do is make itself an indispensable social layer on top of everything else. Doing so essentially makes companies like Microsoft and Google developers on its platform, meaning one day they could need Facebook and its data more than Facebook needs them.

And, search may be the main way that users spent the past decade navigating the web, but Facebook intends for social software — whether through a search engine, the news feed on its own site, or anywhere else — to win in the future.

So making Bing more social just ups the pressure more on Google to figure out social data on its own, whether its own entertainment-driven social destination or social features that improve its other products. Google needs high-quality data on its own, not just to help it personalize search, but to help it displace Facebook’s importance to users and to other developers.

Google’s real fear of Facebook is not that it helps a competitor deliver somewhat better search results, but that Facebook itself displaces Google as the average user’s home page. Once Facebook or Facebook-connected applications are where people engage with the web, Google itself becomes less necessary.