Speaking to the Wall Street Journal yesterday, Yahoo’s next CEO Scott Thompson said he plans to return Yahoo to the status of “one of those great iconic brands” on the web by embracing “innovation and disruptive concepts.”

Nothing’s more innovative or disruptive right now than Big Data, and Thompson should make leveraging the company’s deep experience with Big Data one of his top priorities.

New Yahoo CEO Scott Thompson

Yahoo has been using Hadoop for more than five years and currently boasts the largest Hadoop deployment on the planet. The company uses Hadoop to process huge volumes of click-stream data to match ads with users, detect spam in Yahoo! Mail, and pick top stories for its homepages. Last year, Yahoo spun-out its Hadoop engineering unit into a new commercial Hadoop company called Hortonworks, but it still retains a small army of internal Big Data engineers. And Hortonworks currently only has one customer to support … Yahoo.

Yahoo has been losing online advertising market share to rival Google for years and is now a distant second in that market. In order to get back in the game, Thompson should look for new ways to leverage Yahoo’s Hadoop and Big Data resources to improve the company’s online advertising business, both by improving the algorithms that support Yahoo’s ad business and by making Yahoo’s various properties more compelling for users.

On the second point specifically, Yahoo needs to make its various properties more attractive destinations by becoming more social. Yahoo creates some great content, but the easier it is for users to share that content and engage one another the longer they’ll stay within the Yahoo network and the more likely they are to return (and return often.) That makes Yahoo all the more attractive to potential advertisers looking to target those users.

The company took a step in the right direction last year when it partnered with Facebook to integrate updates and games from the social networking giant with its various properties, but Yahoo needs to develop more of its own social capabilities. Users today expect an interactive experience when they go online and Yahoo doesn’t provide enough of it at the moment.

So how can Big Data help? Like Facebook and LinkedIn, Yahoo should leverage Hadoop to analyze its user-base and connect like-minded users to one another. As Yahoo learns more about connections between users, it will also gradually get better at recommending content and new services based on the preferences of users within similar social circles. Big Data can also help the once mighty web company develop and support new products and services, as well as improve its internal analytics capabilities. Thompson could surely use such capabilities as he makes critical strategic decisions about whether or not to sell off Yahoo’s stakes in Alibaba and Yahoo Japan.

Thompson should be no stranger to Big Data. PayPal, where Thompson was president for the last seven years, made use of Big Data techniques to shore up its fraud detection capabilities under the direction of DJ Patil. In a statement following the announcement of his hire, as first reported by ZDNet, Thompson spoke about the potential of Big Data to help Yahoo company regain its footing:

I feel certain that that wealth of data is going to be exploitable for next-generation products, next-generation experiences, and for super competitive capabilities in the display space in other space databases that are in marketing and advertising. So, a lot to learn, still very early, but my instinct says we’re going to be able to find ways to compete and innovate that the world hasn’t seen yet.

Thompson should waste no time following his instincts. Big Data could prove to be Yahoo’s Big Savior.