A quarterly report from Janrain analyzing social log-in consumer preferences of Web sites and mobile apps for its 365,000 customers suggests that 45% belonged to Facebook in Q3 2013, down from 54% in the year-ago quarter. While Facebook remains dominant, it's losing favor with some consumers who would rather sign in under Google+

Andrew Goodman tells us about five paid-search iterations that marketers should not ignore. He explains how to get them running, such as basic ad group setups, display ad networks, sitelink extensions, and remarketing links for search ads, because there's nothing worse a marketer can do than set up targeted ad groups and forget them. Even the single rookie mistake of going with all broad-match types can knock down return on investments by 80% or more. Read the article here.

Bing has added Pinterest boards alongside alongside regular search results in image search. In April the engine introduced the ability to pin and save results directly from Bing image search. Now searchers can add curated collections from around the Web. It's another example of the integration of search and social. Jon Noronha, Nevin Yang, and Deepak Santhanam provide examples.

Stella Fayman weighs the pros and the cons of outsourcing overseas versus building an in-house search engine marketing team. While lower cost might be among the possible benefits to outsource, quality could become a big issue. As the electronics manufacturing industry learned from years of outsourcing manufacturing and code building to China or India, it's not easy to ensure that the code getting built for the search strategy remains top-notch from thousands of miles away. Read the article here.

How do you accomplish a quality score that lowers cost per click, CPA, and improves ad ranking? Larry Kim maps out the anatomy of a quality score identifying five characteristics, from extensive use of long-tail keywords to strict adherence to best practices. Kim references an unidentified WordStream client for the example, explaining how marketers must analyze the data. The ad on the surface might suggest one thing, but the data will tell another. Read the article here.

In honor of Halloween, Search Engine Journal launched a Festival Fall SEO Recipe for Success Campaign. Judges will select the best Five submissions that will post for votes. The grand prize winner receives a $500 gift card, an SEJ grab bag and a congratulatory post with their recipe for success featured on Search Engine Journal. Bernadette Coleman provides an example such as "1 Crazy Egg, 4 cups fresh content, 3 meaty cuts of Matt," and 8 liters link juice. Sounds fun.

Cyrus Shepard released the full set of data around the 2013 Search Engine Ranking Factors report. The biennial report, supplemented with real-world correlation data -- about 17,000 keyword search results -- aims to help marketers understand the characteristics on pages that search engines consider meaningful. Shepard tells us it provides insights into ranking factors. Read the article here.

Ian Ayres set out to prove or negate Microsoft's BingItOn campaign results by enlisting four Yale Law students to run a similar-size experiment using Microsoft's site. The results revealed that 53% of subjects preferred Google; and 41%, Bing. Six percent of results were tied. He explains this isn't even close to the results Bing claims and wants to know how nearly two to one prefer Bing to Google. He also claims that Microsoft refuses to release the results of its comparison Web site, making matters worse.

Google added support for Hulu Plus on Wednesday. The streaming media fob makes an ordinary television smart. Viewers wanting to add Hulu Plus must download Android or iOS for the service. The company said it will initially launch the system for Android tablets and smartphones, followed by one for Apple iPad. Support for the iPhone will have to wait.

Machine learning company AlchemyAPI has developed an app that uses the Google Glass camera to recognize what a person looks at through the glasses. The image is sent to a database in the cloud. In near real-time, the app searches and compares the image with others in the database, and sends back its "best guess" as to what it sees.