In our last article, we discussed how analytics acceleration helped to deliver faster insights using infrastructure designed for complex analytics. In order to accelerate analytics, IT leaders need to become the architects of the future and employ a new data approach.
IT leaders who are looking to make sure that IT plays a greater role in driving the business forward will architect environments that employ data centric design.
Key Areas for Data Centric Design
Data centric design will help your organization capture the largest volume and variety of data. It will help you efficiently locate data and applications. It will help you make better business decisions faster. Sounds easy enough, but to make this a reality, you’ll need to make sure that your team is ready to focus on three key areas:
- Intelligent data management and placement
- Rapid data ingestion of diverse types and sources
- Moving analytics closer to the data
Where Is My Data? What Is Important?
To architect a data-centric design, you’ll need to consider your business drivers in addition to the data that helps to tell the story of those drivers. Sometimes that data isn’t readily available or available quickly enough to help drive decisions. Sometimes it isn’t even your data, but data from third parties. Consider these cases:
IBM Helps TBV Maximize Ad Revenue
TBV, a Hong Kong-based television broadcaster needed get a more precise ratings forecast to maximize ad revenue. To achieve this, they had to integrate external social media data with over 1 TB of data from its web and mobile applications each month.
The Data Centric Design: IBM Power Systems running DB2 with BLU Acceleration and Storwize v7000
IBM helps LSU use Hadoop to Analyze Genomics Workloads
Louisiana State University uses Hadoop to analyze over 3.2 terabytes of data. By investing in IBM compute and storage solutions, they were able to increase performance by 3x while using 2x fewer notes—a 9x improvement in terms of performance per server.
The Data Centric Design: IBM Data Engine for Hadoop and Spark—Power Systems, Spectrum Scale
Petrol Improves Sales with Historical Customer Data
Petrol and IBM worked to design a data centric solution that helped the European retailer use historical and transactional customer data to uncover new opportunities and improve sales. Petrol reduced data analysis of each new transaction from minutes to seconds. This allowed them to deliver new personalized cross-sell and up-sell promotions at the point of customer contact.
The Data Centric Design: z Systems with IDAA
The Future of Data Centric Design
To start moving toward data centric design means diving into the world of dark data, third party data, and your own data that you capture, but don’t capitalize on.
Consider this: Less than 1% of the data generated every day is mined for valuable insights.
Traditional data capture and analysis models are slowing progress. Data centric design is a new vision for computing that places the processing where the data is stored. Early adopters of this vision are in research laboratories and universities, but increasingly, businesses are tapping into their mountains of unstructured data to help them make business decisions.
IBM – Design for Cognitive Business
Analytics acceleration and data centric design are at the core of IBM’s Cognitive Business concept. In the weeks to come, we’ll talk more about how those foundational elements can help organizations build with collaborative innovation and deliver these solutions through cloud platforms.
Subscribe to our blog to keep up with this story or reach out to us to start building your own cognitive business story.