To compete in the age of the customer, it’s essential to make the most of the data you have access to, whether it’s from internal or external sources. For most organizations, this implies a need to review and challenge existing approaches to how they capture, process, and use data to support decision-making. But it’s important first of all to move beyond a technology-centric view of big data. This is why at Forrester, we define big data as:

The practices and technologies that close the gap between the data available and the ability to turn that data into business insight.

Moving beyond a technology-centric view doesn’t mean, however, that a bottom-up, technology-led approach to big data strategy won’t work. After all, it’s often the case that business executives can’t see the potential of a technology until they’ve seen it in action. A bottom-up approach also provides the opportunity to acquire technical skills, and gain an understanding of what needs to be done to integrate new technologies with existing systems (even if it’s just at the level of getting the data out – often easier said than done). But a pilot project or proof-of-concept demonstrating the “art of the possible” in a business context is different from implementing a Hadoop cluster and expecting the business side to start asking for projects.

Even with the most business-centric approach, there is still the risk that pilot projects are run in isolation, leading to deployments that may address a specific business issue, but without taking the wider business and technology context into consideration. The inevitable result: another data and application silo. Hence the importance of having a big data strategy.  But where to start? And what should be taken into consideration? These are the five key areas that you need to tackle:

  • Find a balance between bottom-up (tech-led) and top-down (business-led) planning. Both approaches have their merits, but neither can ultimately succeed in isolation. If you find that the dialog between business and technology professionals seems to be conducted in mutually incomprehensible jargon, focus on finding a common language. If you can’t, stop your big data initiative – the investment will be wasted.
  • Recognize that there is no single ‘big data’ technology. While Hadoop has a key role to play, big data is about much more than Hadoop (however loosely or narrowly defined). As outlined in Holger Kisker’s earlier blog post, different scenarios require different big data technologies. The exact combination differs between organizations, depending on requirements as well as existing environments. In our report Strategic Planning For Big Data: Getting It Right, we introduce a simple framework that can help you on your way.
  • Big data has many different use cases.While certain topics keep bubbling to the surface (e.g. improving the accuracy of marketing campaigns, augmenting fraud detection, reducing downtime), big data techniques and technologies can be leveraged by any part of the organization. Just like there is no single big data technology, there’s no single big data starting point. Your big data road map needs to reflect not only what your company wants to achieve, but also take into consideration ongoing initiatives and existing technology investments. Just make sure you always start with a question.
  • Make sure that your planning is long-term.  What you don’t want is another set of silos that’s difficult to maintain and expensive to integrate. There will be times when you need to make tactical choices; but it should always be clear how these will impact the long-term strategy, and how any such impact will be dealt with. This is also why it’s important to
  • Put in place an agile, flexible big data platform. You should consider doing so sooner rather than later, to make sure you can cater for different data management and analytics scenarios, including more advanced techniques such as predictive modeling, semantic search and geospatial analytics.

If you want to dig a little deeper, read our report Strategic Planning For Big Data: Getting It Right. Do let us know what you think, and if you have any additional hints and tips, we’d love to hear them. Feel free to email me directly at