To thrive and succeed, online publishers and social media companies must find new sources of revenue beyond the display ad. Display advertising is and will be a necessary part of the overall user experience, but it's not enough.

The real opportunity lies in exploiting "big data" sets acquired from user actions and content to deliver timely and relevant brand experiences and engagement.

At a conference late last year, Fred Wilson of Union Square Ventures talked about how online advertising is "going native."

Google AdWords is the first and best example Wilson mentioned. Paid search is adjacent to organic results, and in fact reinforces the user experience by delivering significant incremental value. Paid search does not interrupt but enhances whatever the user is doing with more information -- and sometimes delivers just the information the viewer seeks.

Another example is Facebook, which is already far advanced in exploiting terabytes of data to create highly targeted, relevant, and socially aware advertising. Facebook's abilities to finely target based on social, personal, and location data are well known. The company has recently moved into native forms of advertising with social-style ads that appear in user's news feeds and combine a message from an advertiser as well as brand-related status updates from users. Facebook plans to expand advertising to mobile as well, according to reports.

In other words, it's a completely new paradigm, built specifically for the web. Banners and even video do little more than attempt to replicate their cousins in print and broadcast media -- and they don't do it nearly as well.

Gaming platforms, social networks, and others will rise or fall with their ability to capture, manage, and generate revenue from the massive data sets their users generate.

For example, consider Twitter. Sponsored tweets are a good example of a form of advertising that is native to the platform. But sponsored tweets will not be enough to support Twitter. Don't be surprised if Twitter explores location-specific sponsorships of products and local merchants. It could enter the local merchant discount or promotion business.

The company has a lot of data about the location of its users, the content they create, and traffic. Imagine if Twitter could notify users of a happy hour just a couple of blocks away and they could get their first drink free. How useful would that be?

Large online companies have the resources and ability to build infrastructure from scratch, but smaller companies will turn to a growing group of "big data" applications that will let them easily see, analyze, and visualize huge amounts of data at scale.

With "big data" analytics, visual sites such as Tumblr and Instagram would be positioned to create contests or sponsored photos or groups of photos that would show people engaged in some fun activity relevant to the sponsor. An airline might do something with exotic vacation photos. A beverage brand could promote snapshots of parties at the beach.

Similarly, on a site such as Prismatic, sponsors could invite users to create newsfeeds around a topic.

One of the challenges companies face is that they often do not know where to start. To take advantage of these potential monetization opportunities, they must excel at bringing together torrents of data from different sources and make sense of it to achieve a business benefit.

For example, check-in data combined with user profiles and point-of-sale data could be used to send some customers special deals. But that's not all: A digital community such as Foodspotting could enable some restaurants, cafes, and bakeries to run contests in which frequent customers could win a prize such as a special meal or an invitation to an exclusive party. Check-in data can also enable brand sponsored, digital real-world scavenger hunts.

Ultimately, "big data" analytics will help advertising become smarter, social, more relevant, and less annoying. It's only natural.

Michael Driscoll is CEO of Metamarkets.

On Twitter? Follow iMedia Connection at @iMediaTweet.