In a blog post published February 2 titled, “Beyond Big Data’s Vs: Fast Data Is More Than Data Velocity,” Forrester analyst Michele Goetz provides her vision of Fast Data. As the title suggests, Goetz does an excellent job at explaining why the absolute speed of data is not the primary goal of Fast Data.
As the company that provides the Fast Data platform, TIBCO is also often confronted with that perception that Fast Data is reserved for a few industries that deal with real-time data. Of course, TIBCO has pioneered real-time messaging on trading floors, telecom operators’ data centers, and high-tech manufacturing plants. Back then—and to this day—these businesses needed real-time information to gain a competitive advantage, ensure quality of user experience or optimize manufacturing processes. Most other businesses could deal with the latency introduced by the processing of data, and its storage in databases and data warehouses. That latency did not impact the services they were providing.
But the digitalization of customers, partners, and their businesses has changed the game. From checking a mobile app to browsing a web interface, and even an in-person visit to a store or a branch, customers access services when they want, where they want it. Customers have access to more data than ever before to compare, review services or items to make an informed consumption decision. Finally, customers know the value of their data and will only provide access to business if they are incentivized—and confident their data is safe. These “Business to Consumer” (B2C) examples explain why analysts mentioned “digital marketing” way before they mentioned “digital business.” There are also many “Business to Business” (B2B) examples that we could have used.
The technologies that enable digital business provide businesses with not only more data to understand their customers better, but also to identify opportunities. Location services on a mobile application are an interesting example. If retail customers that have downloaded the app and agreed to the terms of condition are close to a shop, they can be located and be sent an offer. They will be incentivized to enter the store, but only if that offer is sent while the customer is still in its vicinity. The window of opportunity is short—just a few minutes. And customers will enter the shop if the offer is relevant. This offer needs to be based on the customers’ preferences or their shopping habits. The retailer can also find additional benefits, such as proposing offers on items that are slow to sell or whose inventory is too high.
For this retailer, Fast Data is the ability to access the right data necessary to make the decision and take action to send that offer, while the customer can still get to the store. Of course, the data in motion is the application data that will locate the customer. But the customer profiles, its preferences and habits, as well as data about the offer, need to be available at that moment to be correlated for the offer to be sent as soon as possible—and in the next minute for the notification to reach the customer while he can still walk to the shop. Technologies such as Complex Event Processing (CEP) allow the definition of such correlations, combined with rules to define the right actions. But integration technologies are also key to make sure the initial event (the location of the customer) is identified as soon as possible, and that the relevant systems or persons can be notified. The quality of data is also essential, as it will impact directly the quality of the decision. Master Data Management (MDM) technologies ensure businesses have a trusted version of their customer data on which to base such decisions.
But if we look at the way this decision is modeled, we then uncover a new side of Fast Data. Customer tastes change… so does fashion. Retailers have access to all of their transaction data, as well as a wealth of social data. Predictive analytics (provided by solutions such as TIBCO Enterprise Runtime for R) on this data allows the creation of models predicting the customer population likely to be interested in an offer. This information is an important element in the decision to send an offer, and which offer, to the customer that has been located close to the shop. By making this information available as soon as business analysts have identified it, retailers are differentiated by the relevance of their offers. Providing these same business analysts with instant awareness allows them to assess the relevance of their model and the way to improve them.
Fast Data, in most contexts, is the ability for the right person to access the right data, to make the right decision. As Forrester analyst Michele Goetz explains, Fast Data is not about the fastest data. Whether your window of opportunity lasts hours, minutes, seconds or milliseconds, it is the ability to identify the opportunity, and be able to change its outcome. Fast Data is about making decisions at the speed of your business.