2014: The Year of Customer Intelligence
Big data gave marketers the ability to centralize data. For many marketers (though rare with vendors), most deployments weren't of the Tupperware container-based "push anything you want to us and we'll store it" variety. While that can be used for segmentation, the truly innovative piece to this puzzle is contextualizing every piece of data, be it behavioral, purchases, returns, demographics or applied learning. What that means is that each piece of info stored in your "big data" is "known and expected." Why is that important? Read on.
Large-scale machine learning
If your big data platform understands what your data is, it can ask questions about your data and store the answers. I know, it sounds very Terminator-esque, but it's not that crazy. One of the most basic questions is "how engaged is a specific customer currently compared to the past?" When the system arrives at an answer and stores it, it's learned something.
That said, there are vendors whose primary focus is building machine learning. Some of these companies are building vast libraries of machine learning and enabling more complex questions to be asked and answered, such as "who out of all my customers is most likely to leave my brand, and are they important to save?" While the machine learning itself is rife with potential, it means nothing without the ability to act in place.
2013 was the year everyone started talking about omnichannel marketing. From email marketing vendors to product recommendation providers, everyone was clamoring to attach themselves to this concept. As a quick reminder of what omnichannel marketing really means (since the true definition is often misunderstood), omnichannel marketing means every marketing channel, including display ads, email, on-site messaging, print and social, are all delivering the same message, offer or product to consumers. That said, most of these vendors are dealing with incredibly small data sets and almost no customer-centric machine learning.