The argument for jumping into the fray with big data goes something like this:
- there is an increase in the volume, velocity, and variety of data flowing around companies that is driven largely by social media and that is increasingly comprised of unstructured data – tweets, likes, etc.
- companies that can harvest and mine big data will out-perform companies that don't.
As with any business or technology innovation, the proponents can always marshall case studies to support the value proposition. That being said, what is the right big data strategy for your company?
For our purposes here, let's make a distinction advanced by other industry observers and practitioners – that “big data” is mostly about the large volume of unstructured data generated via social media and Internet behavior. Sure there are large volumes of stuctured data – transactional data and reference data stored in ERP systems and other enterprise applications. That being said, leveraging sturctured data has been the focus of business intelligence, analytics, and data warehousing for 15+ years and is well-covered elsewhere.
If big data is unstructured data, it is useless for business purposes unless it can be organized and leveraged to either increase revenues, reduce costs, or both. Another way to put it is to say that unless big data can be converted to business intelligence and/or analytics, it will be of little business use. Accordingly, I submit that formulating a big data strategy requires the same approach as formulating a business intelligence/analytics strategy. We need to answer the following:
- What information do we need?
- For what kinds of analyses?
- To improve which business decisions?
- Via what changes to which business processes?
In the world of big data, this translates to a more basic question: how will analyzing unstructured social media data and web site behavior data help improve our business results? Will it help in marketing, sales, and customer service? Will it help our supply chain organization? Will it help with financial management? Will it help with operations management? There are case studies to support a “yes” answer in some of these cases, but each company needs to consider the questions from their own perspective.
More broadly, there are some other strategic questions to consider.
- Are business intelligence and analytics strategically important in your industry and/or to your company? Generally, the more complex your business, the more important it becomes to leverage business intelligence and analytics to compete effectively, serve your customers well, and optimize costs in relation to revenues. As some business move more and more in the direction of being digital businesses, unstructured big data will become more strategically important.
- How experienced and successful is your company at leveraging regular stuctured data for business intelligence and analytics? My experience with leading companies in financial services, manufacturing, distribution, retailing, and other industries suggest that some companies are not even ready to leverage small data or medium data – or whatever you want to call the gigabytes or terabytes of structured data most companies have. The are at least 5 Barriers to BI Success that impede companies as the strive to leverage business intelligence and analytics to improve results. If your company isn't good at the basics, it may be a stretch to jump into big data in a big way – and perhaps an R&D pilot would be a useful strategy.
- How complex is your business – and how digital? The more customers, products, channels, and geographic areas your company needs to manage, the more complex your business. Business intelligence and analytics are great for coping with complexity and optimizing the business. Big data can be very useful for improving marketing, sales, and customer service in businesses that heavily digital – think Amazon, online banking, and so forth.
- There is no magic answer for what the optimal big data strategy is. On the other hand, there are structured methods for getting to a practical answer for your company. Considering the questions we've posed are a good start toward formulating a big data strategy that is clearly focused on an important business improvement opportunity – one that will yieled a reasonable ROI.
by Steve Williams, President
Copyright 2012 DecisioniPath Consulting