After reading a recent LinkedIn post by Jeff Haden, "Want to Achieve Lifelong Success? An Army Ranger Says You Need This 1 Trait the Most", (spoiler alert: It's adaptability) something occurred to me. One of the reasons I enjoy solving business problems with analytics is that analytics is all about being adaptive and causing change (i.e. growth).
When you're looking to use analytics, don't look for a black box approach to solving a problem, because once you implement that solution, guess what happens? Life or situations change and your analytics need to adapt and change quickly in order for you to continue to succeed.
What can you do about it? Here are four adaptability attributes for analytics success, and what they mean from a business and technology perspective.
- Business: The ability to try out different ideas without having to switch to other tools, languages, or physical environments.
- Technology: The ability to deploy analytical software capabilities quickly without the hindrance of securing new hardware resources in advance.
- Business: No more failed jobs due to "unknown system issues." The process, job, or model continues to work seamlessly without interruption.
- Technology: Having an architecture designed for high-availability means there are fewer fire-drills, and upgrades occur with minimal disruption.
- Business: The ability to fail fast, try alternatives, and evaluate the results quickly so more ideas can be researched, and the ability to move new results into production now instead of later.
- Technology: Having an architecture that supports the requests of all users in a timely manner, and the ability to deploy insights from models into production.
- Business: Users are no longer limited in their data exploration and modeling by the size of their data. As the problem grows the ability to process grows with it.
- Technology: Since analytical workloads are variable the environment needed to process these workloads should be able to grow and shrink as needed. For example, a scalable environment will burst to additional compute nodes and/or spill to disk if available memory is short.
Your next step? Learn how analytical success starts with the right analytics platform.