Author Tyler Allbritton, Director of Data Analytics for CenturyLink
Many years into the hype cycle, Big Data is still an illusive dream for most enterprises.
Many of the claims of promise and opportunity are very true: a simple Google search (from the ultimate data monetization company) will point you to a diverse set of successes from staid industries like insurance (MetLife Wall) to nimble startups shaking things up (Uber).
Still, most enterprise data remains “un-tapped”; in many cases data remains completely unused and data monetization is a rarity. Can the promises of Big Data be met? Absolutely, but the size, complexity, speed and structure of the data have nothing to do with it. The problem is not that enterprises don’t know what to do with Big Data, the problem is that they don’t know what to do with data (period).
Over the last 20 years, the IT data industry (data warehousing, business intelligence, etc.) has been on a centralization and conformity mission.
We spent a lot of time and money and largely succeeded at our mission, but missed an important hidden element: business context. We took orders for data and delivered them without understanding why. Many reports were delivered to specifications, but they were never correct; there was always the need for the next report variation. We lived with this and told ourselves that data was “just different.” However, the lack of context – not understanding the problems our business partners were trying to solve – has always been the failure and Big Data is not going to fix it.
Is there a “Big Data” Hidden Element – Business Context?
A data system in and of itself does not solve anything. It makes data available to the enterprise masses to achieve efficiency gains (one stop shop, reliability, etc.), but this is not earth shattering in terms of value add. Those data were always somewhere and someone could always get them. Until you understand the context – why the data are needed – it will be very difficult for you to ensure your data solution will add value at the end of the day.
Thankfully, the process of defining a business problem aligns wonderfully with data. Identifying the true owner of the problem, the need, and the desired successful outcome at a detailed level is non-technical and non-trivial, but fundamental to success. Treat business problems and potential solutions as hypotheses. This drives one to prove with facts and data, not just intuit, the existence of something. In many situations this approach can highlight that a problem may not be the biggest problem or even a significant problem at all. Once the problem (or hypothesis of a problem) is understood, identifying the appropriate data to describe it is not a herculean task. Empirically proving the problem is refreshing and empowering for all involved.
So ask yourself – is your goal to implement “Big Data” Solutions or to find Solutions to Big Problems?
An ancillary benefit of this approach is significantly narrowed scope providing focus and common understanding with business partners. If new, Big Data are required, it will be apparent – there will be no “searching for the right Big Data use case.” In many cases for most enterprises though, Big Data will not be required. And guess what, your business partners won’t care because you helped them solve a business problem. They don’t want or need Big Data; they need to solve Big Problems.
Make your business partners work a bit and explain their business problems to you. If you just focus on building it bigger, even if they do come, they won’t know what to do when they get there.