Promise of Data Analytics
By John Kochavatr, Chief Information Officer & Digital Leader, Suez
Data and analytics provide a lot of opportunity for businesses these days, offering economically viable ways to generate new algorithms, enable scalability, and remove barriers so businesses can be more nimble and adaptable than ever before.
New capabilities lower the barriers to entry for many companies in the analytics space, and this serves as both an opportunity and a threat for many business models, including ours. It serves as an opportunity for businesses to move quickly.
These new capabilities can be a threat when non-traditional competitors are able to bring new solutions at lower costs to your customer base. It is certainly interesting times in this digital world.
Some of the challenges while deploying Data Analytics
The concept and promise of analytics is very exciting. However, getting it right is difficult for most of us; and not for lack of effort. One of the first and hardest questions is, “How do we define success in this space or on investments that we make around analytics?” Business are often finding that, while they have tons of data, the taxonomy of that data isn’t quite right, because it was never intended to be used in the way businesses want to use it today. Reconciling this and understanding what success looks like is imperative to make the best use of resources.
Steps that Data Analytics can take to foster innovation and/or growth
To use Data Analytics in ways that foster innovation and/or growth, there are a number of steps that we take. First, and most importantly, define success by identifying specific outcomes and estimating returns on investment from these outcomes. It’s easy for these types of programs to become science projects that churn through investment dollars, if you are not careful.
The concept and promise of analytics is very exciting
If you don’t do this work ahead of the game, you may end up with suboptimal results and be disappointed.
Then, create a high-level roadmap on how you can achieve some of these outcomes. Your map should include a multitude of parameters, including number of resources, budget, and timeframes. Understanding how you can get from point A to point B will help you, your team, and your sponsors connect the dots on the value of your initiative.
And finally, experiment with a “fail-fast” mentality. In some cases, it’s more economical to try something and fail rather than get caught in analysis paralysis. If nothing else, you will most certainly learn a ton from this approach and gather data points on what does and/or doesn’t work.
The unique lessons learned with the rich experience of managing IT organization
With any program, data analytics being no exception, business must focus time and effort on driving business and/or customer outcomes, and these outcomes need to be economically measurable. While programs like data analytics are very exciting, most of us are measured on results that translate into helping our enterprise economically.
The role of the IT team is an ever-changing one. Gone are the days of pursuing either an IT track or an operations track. I’ve always believed (and still do) that these roles are converging and the best of us will have expertise in both. Strong operation leaders have strong IT experience, and strong IT leaders have a strong understanding of operations. Use your experience and expertise to drive meaningful outcomes for your enterprise, but make sure whatever benefits you strive for are ones that hit the ledger with hard results.
A wish list of solutions to look forward to
Over the years, businesses like ours have collected data intended for certain objectives. As we look to develop more advanced analytics across different assets and operations, we are repurposing this data and finding ways to connect disparate and unstructured data. I would love to see a solution that helps us more easily stitch this data together.
For example, we have millions of data points in our lab systems that test water and oil quality, which serve as a rich source of information about assets we treat globally. When these systems and processes were developed, the output was simple: log the sample results for the requestor. Now, we want to develop analytics using this data and asset performance data to create digital twins and new algorithms for asset performance. The data structure and taxonomy weren’t designed to integrate with asset performance data, so we are working diligently to rearchitect how we bring these data sets together. This is already helping us drive better outcomes for customer with improved treatment methods and optimized asset operations, which leads to lower costs and faster cycle times on issue resolution. We could not be more excited about where we are going.