It seems like everywhere I turn these days I’m seeing something new about the need for big data and the associated analytics that necessarily goes along with it. Big data refers to the collection of massive amounts of data and the extraction of actionable information from that data, often in real time. Although big data has been around from a technology standpoint for at least 10 years (even longer if you count its predecessors, business intelligence and data warehousing), its time has certainly come. World markets are evolving and changing so rapidly that big data analytics has moved from a “nice-to-have” to a “must-have.”
Apache Hadoop™ seems to have emerged as the de facto open-source platform for processing and storing massive amounts of data, with support and services coming from a variety of fast-growing companies like Hortonworks, Cloudera , and MongoDB (to name a few). These are what I like to think of as the fun companies—start-ups experiencing rapid growth. Of course, all of the traditional large IT companies like IBM, Intel, and Oracle have been in this field for quite some time, along with most major universities with training programs.
So, why is big data getting all of this attention, and what will the impact on projects be?
To answer these questions, let’s see how big data analytics might play out in one hypothetical scenario. A recent article in Fortune Magazine, “The Billion-Dollar Bourbon Boom,” provides a good starting point for our scenario. It turns out the bourbon industry is experiencing a boom. Distillers can’t keep up with demand. Liquor stores can’t keep high-end bourbons on the shelves. A great problem to have—or is it?
The last time demand was this high was in the ‘50s and ‘60s. Distillers added capacity like crazy, and small distilleries sprouted up all over the place. Then came the ‘70s, and bourbon sales dropped dramatically. Distilleries went out of business. They had no data explaining the trend, and so didn’t see the change coming. What happened is that consumers started drinking other beverages like wine and, later, vodka. Sales for bourbon remained flat worldwide until recently. The bourbon industry doesn’t really know why it has grown so much recently, nor how long the boom will last. Given that the product development cycle can span four years from distillation to sale, not knowing can be very risky, just like it was in the ‘70s.
So, how can big data analytics play out in this type of scenario? Let’s imagine a company in this industry can stream all kinds of consumer information from bars, restaurants, and liquor stores around the world. Perhaps they see trending data that show bourbon sales have been growing rapidly in a small country (say Jamaica). Interestingly, they also see coconut milk trending up. Is there a correlation? Enter Project #1, which will drill down further into the data to determine whether these parallel trends signify the emergence of a new mixed drink. Assuming this is the case, the company may send a marketing research team to Jamaica for further analysis. If the analysis yields promising results, Project #2 could involve productizing the drink for this market and beginning a marketing campaign there. Project #3 focuses on market testing for this new product in other locations. Project #4 could possibly be about rebranding this new drink as a separate product line. Project #5 might investigate whether a lower-grade bourbon with a shorter development cycle could be used for this new product, since the bourbon is mixed rather than consumed straight.
Generally speaking, these types of projects are composed of small teams and run for fewer than six months each. While the above scenario is hypothetical, big data has the potential to change the outcomes on many projects. By their very nature, these types of projects can rapidly create new products, mitigate risk, and serve new markets. Whereas companies in the ‘60s and ‘70s invested in capacity without knowing what future demand would be, modern companies, through the use of big data analytics, can build capacity that is incremental, geared toward multiple products, and targeted to specific markets.
Risk is greatly reduced, and at the same time, modern-day industries can sustain growth with a portfolio of products rather than a single product. An evolving market-driven portfolio of products can be maintained with the right amount of meaningful market data, and big data analytics can provide that data.
Let us know how big data is affecting your projects!
Apache Hadoop is a trademark of the Apache Software Foundation in the United States and/or other countries.