Big Data On Its Own Does Not Have A Tremendous Value - Interview With Daniel Kirsch
Yana Prokopets 12 July 2016
Daniel Kirsch- Principal Analyst and Vice President of Hurwitz & Associates, a boutique strategy consulting, market research and analyst ﬁrm that focuses on how technology solutions solve real world customer problems. Hurwitz research concentrates on disruptive technologies, such as Cloud, Big Data, Analytics, Social and Collaborative Business, Service Oriented Architecture and Web 2.0, Service Management, and Information Management.
Big data has been a buzz word for a couple of years already. How does this technology trend transform the business today?
Big data on its own does not have a tremendous value. I have seen many clients struggle and fail with big data projects. The reason for these struggles have been that the project was not tightly aligned with business challenges or opportunities – getting insights into your data is interesting, but when not tied to business objectives the value can be lost.
When Big Data is combined with data management, governance, and advanced analytics the business impact can be tremendous. I have seen companies unlock value insights from the data that they already have but are not analyzing. Take for instance a retailer – many have huge amounts of data on their products and customers however few are analyzing this to offer customized offers or anticipate the trends for next season. Some industries where we are seeing a huge impact on big data and analytics are in financial services, healthcare and oil and gas. The commonality between these industries is that small improvements based on data can lead to millions of dollars of revenue, massive savings or quite literally the lives of patients. Big data brings big risks. What are security concerns to consider when it comes to big data management?
It is alarming how many organizations ignore data security and privacy when exploring big data projects. Basic data masking procedures can be disregarded when doing an analysis of customer data. The following are some considerations for big data and security:
- Mask personally identifiable information - Know who in the organization is touching sensitive data - Baseline the usage patterns of big data so that if there is an anomaly you can detect it – example: if data isn’t typically being accessed at 2am or from a foreign nation you want to know when it is - Develop privacy and security best practices within your organization that must be adhered to whenever embarking on a new big data project
What database management systems would you recommend to our readers who want to leverage big data?
The database management platform really depends on the use case. Some considerations: - How much data are you analyzing - What types of data – structure, unstructured, a mix of both? - Do you need the results in real time or can you wait for the analysis? - Will new data be fed into the environment weekly, daily, hourly or in real time? - How does your enterprise currently manage data and can it be integrated into a big data project rather than moving the data? - Is the data and workload appropriate for a public cloud?
Judith Hurwitz dedicate one part of her book Big Data for Dummies to big data analytics. Can you give some advice on how to apply different types of analytics to business problems?
For years marketers have been performing customer analytics to understand customer churn, up-sale / cross-sale opportunities and customer value. Other types of analytics and industries: - Financial Services – Credit risk and fraud - Manufacturing – supply chain, predictive maintenance for equipment - Workforce / talent management – Employee productivity and value, compensation analytics - Healthcare – predictive analytics on patient outcomes (for example patient hospital readmissions) - Oil & Gas – Analysis of both structures and unstructured data to understand the viability of a proposed drilling site or to find new sites. Also big users of predictive maintenance for equipment
What books or other resources can you recommend to our readers who'd like to learn more about big data?
I’d of course recommend Big Data for Dummies by Judith Hurwitz, Alan Nugent, Fern Halper and Marcia Kaufman. Because the area is changing so rapidly I find that youtube videos and web-articles are a great source of information.