Bulk Up on Big Data from Your Mainframe, Part 2
In Part 1, I talked about the trend for organizations are supplementing big data initiatives with big iron data and how Radixx International leveraged z/OS data and IBM’s SoftLayer cloud to speed and secure core processes for its e-commerce sales and distribution service for more than forty airlines. Now, let’s look at some more use cases.
First Party Data
The most important data sources belong to your company. Some refer to this as “first party data.” For most organizations, first party data means revisiting your own applications in a new light.
As Merck’s CIO Clark Golestani put it, “. . . it was time for us to think about ourselves as a data-driven company and become an analytics powerhouse.”
Where to start? The data sources within enterprises begin with the obvious:
- Make archived data live to enable analytics, such as historical trends and customer profiling.
- Instrument existing applications supporting products and services to collect more data: more often, on more people, and at greater granularity.
- Leverage access to data within sister divisions, operating groups, or sister companies.
- Deconstruct existing transactions to give them greater usefulness. For example, when marketing gets a lead, save everything known about the context of the lead – down to the clickstreams and audio recordings of voice calls.
Use Cases Demonstrate how Organizations are Bulking Up Big Data Respositories — with Mainframe Data
Join the Third Party . . . Party
Now that you’ve pulled in all this enterprise data, and your data science people are getting schooled in the latest analytical tools, you’re ready to start bulking up even more – with other data sources. Third party data can come from diverse sources: supply chain partners, retailers, social media, perhaps even satellite imagery.
Boston Consulting Group explained how a biopharma company was able to use big data to design more effective clinical trials. The patient selection process began in the usual way, but then the firm reached out to electronic health records from a provider partner and combined this with “de novo” data from other empirical sources. The integration of EHR big data from partners with first party data typifies the use of third party data sources for new insights.
Data is ubiquitous – but sometimes it can be hard to see the forest for the trees. Many companies . . . believe they have to collect their own data to see benefits from big data analytics, but it’s simply not true.
Recently Bernard Marr directed those remarks at Forbes readers. Marr goes on to list 35 free third party data sources that he believes are representative of many more in niche industries. They include data from the US Census, the EU, UK, Canada Open Data, the CIA World Factbook, UNICEF – as well as commercial sources such as Facebook and Amazon.
Third party sources should fill out the Variety element in your Volume – Velocity – Variety Big Data portfolio.
Bulking up for Decision Support
The ultimate purpose of this bulking up is to power decision support to improve profitability and product quality. That’s the shiny upside.
But it can be just as important to powering heads-down agility. That’s what PTC Kepware’s oil and gas clients needed when oil prices plunged precipitously last year.
A proven way to supplement data from archives, third parties and z/OS itself is the Syncsort ZPSaver Suite, which leverages IBM’s economical zIIP hardware to speed copy and sorting operations.
You can also use Syncsort’s Ironstream to deliver z/OS machine data in real-time to Splunk Enterprise® and Splunk Cloud™ in real-time to integrate with diverse enterprise-wide data sources for analysis