Last month, we hosted a Data and Analytics Fireside Chat Webinar. Darren O’Neill, senior director of Manufacturer Data Solutions at McKesson, and Sirius’ VP of Data & Analytics Solutions Joe Grant Bluechel, answered questions submitted by the participants.
This unique Q&A session helped attendees gain a deeper knowledge about their data to help improve their analytics plan and advance critical strategies. Here are some of the highlights.
How are companies balancing business demand with necessary engineering?
This question points out another more relevant question, and that is, What is the business need vs. what is IT recommending? While both are important, we’re starting to rebalance and rethink how the architecture or engineering affect the business, and how the business affects the architecture. Look at it from a business perspective: what do we need to do to produce the results we want? We need to look for ways to prove value quickly; not always to shoot for best possible solutions, but to iterate quickly and pick up the velocity. Once we prove out the value, the solution can be designed so it is hardened and scaled. So, we kind of changed the conversation to have the business demands lead the conversation.
In my organization, the number-one problem that we’re being asked to solve is around supporting this whole concept around bi-modal IT. It’s important that clients still realize that the rigors and governance around having a core competency and a mode of operation that is enterprise-class and highly scrutinized is important, but you need the second mode of operation around data discovery, business agility, and failing fast. A lot of times we want to get data into the hands of our data analyst so they can see if there’s repeatable analytic value there. If there’s not, they need to move on fast.
How are industry data models helping organizations get targeted in their business?
They are seen as solution accelerators and starting points, not necessarily be-all, end-all, off-the-shelf solutions. The reason for that is a couple of things. Number one, as a business, are you willing to change your business process and policies to match the solution?… The second aspect of that is that the ownership should fall on the organization to study where the solutions came from. Sirius is very well versed in what solutions make sense for our clients… But the bottom line is that these solutions are there to help improve the velocity to implementation.
Other questions addressed included:
• Outside of unstructured data, is there really an advantage to using Hadoop? And are data lakes real?
• What use cases and/or patterns are easiest to implement in the cloud?
• Do you run advanced analytics models on streaming data sets?
• Should our data integration strategy utilize a virtual mediated schema to integrate various sources?
• What do you envision as the best org structure for analytics, and how much does corporate culture play into determining the best structure?
• How do you bring predictive analytics solutions into the hands of skeptical, non-technical users?
• As much as we try to engage the business in self-service, it does not seem to be where the business want to go. Is self-service really an adopted practice in most organizations?