Why is Fast Data, the New Big Data, Gaining So Much Importance for Businesses?
Enterprises need a technology stack capable of ingesting and analyzing fast incoming data, which has the ability to instantaneously act on that data by analyzing the live streams in the context of insights extracted from historical big data stores.
Data holds immense importance for all enterprises irrespective of the industry or domain because it provides powerful insights for opportunities, responses to market demands, decision making and much more. It has the maximum value, the moment it arrives, and that usefulness gradually diminishes while it stays in storage. In other words, the data in storage loses value every second, since the ability to notify or act as things are happening is lost and an organization that is unable to act on incoming data instantly, is getting incompetent and obsolete. In a literal sense, a majority of enterprises that consider themselves data-driven actually aren’t.
Big data is data at rest, that is often created by data which is generated at incredible velocity as well as volume, such as financial ticker data, click-stream data, log aggregation, or sensor data. Around a decade ago, it was almost impossible to think of analyzing petabytes of historical data. Today, open source technologies like Hadoop reimagined how to efficiently process petabytes upon petabytes of data using commodity and virtualized hardware, making this capability available to developers everywhere at a lower cost. Consequently, the field of big data emerged.
Big data is not just big, it’s fast too. However, the benefits of big data are lost if fresh, fast-moving data with ongoing purpose is dumped into HDFS, an analytic RDBMS, or even flat files. Processing big data tells you what happened at some point in the past, a week ago, a month ago, last quarter. It provides historical context, making it possible to say, for example, “This investment in an event, last quarter would have made us N% more productive, so let’s do it going forward.”
Fast data, in contrast, is all about the moment. It is data in motion and it needs technology that has the ability to analyze, decide, and act on, that is, offer recommendations, make decisions, or take other actions, as fast as data arrives, typically in milliseconds. Acting on data as it arrives has been thought of as expensive and impractical if not impossible, especially on commodity hardware. Just like the value in big data, the value in fast data is being unlocked with the reimagined implementation of message queues and streaming systems such as open source Apache Kafka and Apache Storm, and the reimagined implementation of databases with the introduction of open source NoSQL and NewSQL offerings.
Fast data deals with extracting insight and creating business value through analytics and processing on data in motion instantaneously as it streams into the enterprise. The data can be from any source - customers environment, IT systems, financial sources, security infrastructure, IoT sensors, logistics, manufacturing units, field operations, and so on. The gap between resident and fast data represents unused value for any business, and the ability to harness that value will directly drive business success. So the faster you can act on the incoming data, the more value you can extract.
While organizations have spent the past decade making massive data investments to outperform competitors, the maximum amount of that investment is focused on data warehousing, data lakes and other systems for using resident big data. A decade ago, a relatively lower investment could gain a huge competitive advantage. Today, companies are investing large amounts to pursue increasingly marginal returns. Winning organizations have started differentiating now and will continue to do so in the future, basis the ability to look and act on data faster.
Definitely, some industries and sectors have already taken the step to embrace data in motion. For example, in financial services, everything happens in real time, and being able to assimilate as well as act on data rapidly is not only an indispensable part of the game, it’s a prerequisite to establishing a competitive edge. In fact, almost any online transaction across financial services, retail, telecommunication, travel and hospitality, social or streaming media, needs to utilize fast data for greater success.
Let’s take the example of implementation of live interaction with prospects or customers to improve conversion and satisfaction, automated systems that rapidly adjust recommendations based on data analytics or even the development of data-driven products that create new revenue opportunities for businesses. Each of these innovations has been applied in some way by Amazon. The company stays atop in retail not because it has a better understanding of each of the products it’s selling. Rather, according to tech giant Google, a critical factor is Amazon’s ability to utilize fast data to drive customer success.
Having said that, most of the modern day enterprises are just gearing up to consider fast data, implying smaller investments can produce exponentially higher returns in terms of competitive advantage and differentiation. In fact, fast data processing is getting tagged as a prerequisite for some of technology’s most exciting advances. Building data-driven thinking into all processes will help evolve a data-driven culture in an organization. As the urgency of superior customer experience becomes a concern for everyone across the enterprise, training, and education, along with access to tools and platforms, will be essential for driving employee engagement as well as foster respect for leadership.
Consider machine learning and artificial intelligence (ML and AI) as an example. Most analysts and data scientists consider the application of ML and AI to automation, predictive analytics, security and other use cases to be potentially transformative for their industry. Research labs globally are working enthusiastically on building and advancing the underlying data models. Yet extracting value by putting those models into production has become the major challenge.
Automation, as an example, requires the right input into the right model at exactly the right time and the ability to learn, i.e., adjust the model, as new data becomes available. It’s that ability of modern automated systems to make quick adjustments based on data analytics, the machine learning that distinguishes them from the rudimentary robotic assembly lines of past years. Fast data is all about that ability to gather, join and transform data in motion, making it fundamental to machine learning and artificial intelligence.
The internet of things (IoT) can be another example, that is predicated on the ability to gather and, more importantly, act on the tremendous volume and velocity of data from connected devices streaming into a business enterprise. IoT use cases such as adaptive maintenance, security monitoring and governance, predictive repair and process optimization, all need the propinquity of fast data. It’s not only technological advancements, but also the opportunity to explore entirely new business strategies or concepts, that is exciting about the world of fast data. Those enterprises that recognize and embrace this ability to act on data as it arrives are clearly going to gain a competitive advantage.