Like any new technology, the hype around big data is starting to fade as the hard work of implementation begins for many organizations.
Big data may be the hot term, but it’s the resulting analytics that are driving the investment. Whether you’re far down the implementation path or just setting out, you already know how difficult delivering those analytics can be.
I’d like to share a framework of eight elements that are essential for the success of any big data project.
Mastering these elements will help you deliver the competitive advantage that led to your big data investment in the first place.
1. IT’S MORE THAN TECHNOLOGY
The first element, of course, is the underlying technology. This is where most IT project teams naturally gravitate, and it’s essential, no doubt. But it’s only one small piece of the analytics puzzle.
2. DATA GOVERNANCE
The old expression “garbage in, garbage out” absolutely applies here.
The larger and more complex your data environment is, the more you need a strong Master Data Governance policy in place to ensure the information coming into the implementation is clean, and truly compares information “apples to apples”.
3. PROCESS GOVERNANCE
If your processes for collecting data aren’t consistent and complete across your businesses and regions, you’re analytics won’t be accurate.
Let me give you an example from our business: If one Avnet business books inbound inventory the moment it arrives, and another books it at the point that it’s stocked, there’s no way our cycle time reporting models could be accurate.
4. STANDARDIZED REPORTING
At Avnet, we had thousands of different reports being created in our old business intelligence system.
A big data implementation is an excellent opportunity to develop a standard set of reports that delivers one version of the truth, aligned across all business leaders.
5. HIRE THE RIGHT TALENT
Once the implementation is complete, you’ll need true quantitative data scientists to properly create the predictive scenarios and analytics models you need.
These individuals may already exist in some organizations, but many others will have to go out and hire that skill set.
6. GET I.T. OUT OF THE WAY
You’ll want to provide a select group of business analysts with the tools they need to run ad hoc reports and “what-if” scenarios for your leadership team.
All IT needs to do in that case is deliver the right data into the environment where these user-driven tools operate, and let the analysts do the rest.
Keeping IT’s role minimized once the implementation is done leads to faster reporting and greater agility across the organization.
7. IDENTIFY DELIVERY CHANNELS
Once your reports and models are ready to run, how will that information be conveyed?
Do your business leaders want information pushed to them through email, or do they prefer to pull the reports on demand?
Ensuring that the right data is conveyed in the right way to the right people at the right time is vital to the success of your analytics implementation.
8. DON’T LEAVE ROOM FOR INTERPRETATION
Analytics alone are useful, but providing unique intelligence that helps business leaders make the right decisions is the ultimate end goal for any big data investment.
Be sure to support your leaders with the assistance they need to help them understand the analytics before they make critical business decisions based on it.
DEGREES OF COMPETENCY
While we at Avnet believe that every big data or analytics implementation has to have each of these eight elements in order to be successful, consider it a journey rather than a box to check in the project.
Each element has many degrees of competency within it. The more competent you are in each element, the more accurate your analytics will be. And the more accurate your analytics are, the more successful your big data implementation will be as a whole.Tags: Avnet, Avnet CIO, big data, big data implementation, Big Data in the Enterprise, big data success, data analytics, data analytics implementation, data analytics secrets, enterprise IT, IT Best Practices
Categorised in: General
This post was written by Steve Phillips