This article is for you if you sometimes read the last chapter first, like the concept behind the Easy Button from Staples, or love drawing the MONOPOLY card that says "advance to GO and collect $200."
The Utility Analytics Institute's analytics value curve, depicts a step-by-step analytics journey where value increases each step of the way. It represents the analytics building blocks that ultimately lead to business transformation. However, the path to advanced analytics does not need to be a serial process. The great news is that you don't need to wait! You can realize the substantial returns possible from predictive models and other advanced analytics often in less than six months and for relatively modest investments.
In the early stages of the analytics value curve, your attention is focused on understanding your data and answering a myriad of questions such as:
- What data do we have?
- How accurate is the data?
- What data sources are involved, both internal and external?
- How will we integrate our data?
- Do data definitions vary?
- How will we validate and transform our data into standard formats and eliminate duplicates?
- How will we manage data updates?
- How will we store and secure the data?
- What tools will we use to access the data?
All of these questions and more will need to be answered to build a solid data foundation. This foundation will provide access to your data and help generate standard reports about what happened in the past. Additional tools and capabilities will deliver ad hoc reports that will answer questions like when, how many, where, and how often. A more robust toolset will enable you to create queries and drill down to discover the root cause of a particular problem, set alerts to trigger appropriate actions, and create dashboards to meet the needs of specific user communities. All of this is part of the evolution to becoming a data-driven organization.
Access to and use of real time -- or near real-time -- data, statistical and quantitative analytics, forecasting, analytics models (e.g. explanatory models, predictive models, decision models) will take it to the next level and dramatically increase the value realized (note the hockey stick shape of the analytics value curve). You will be able to answer questions like:
- Why is this happening?
- What will be the result if these trends continue?
- What will happen next?
- What is the optimal result -- how can we transform our business?
You can use predictive models to target customers for your new demand response program, produce statistically valid risk scores to improve collection results, or to schedule and optimally size transformer replacements. Substantial returns can be achieved in a relatively short timeframe often for modest investments. Development of an enterprise-wide analytics foundation can be done in parallel.
This approach is consistent with the idea of achieving those quick wins we often hear about. Prioritize and carefully select those projects likely to deliver the most significant results. Pursue these initiatives as part of a comprehensive plan and avoid one-off investments. Most utilities will find it best to work with a vendor partner able to assist in the planning efforts, offer unique expertise, or provide a specific analytics toolset.
Why rely on a weighted point system and risk scores that represent past behavior when statistically valid predictive models can predict future behavior and improve debt management? Gaining experience early on with advanced analytics will provide invaluable experience, enable you to refine your enterprise data plan accordingly, and deliver the early returns that will garner the support needed for future analytics investments. Attend Utility Analytics Week to learn more!
Kim Gaddy is a consultant and a Utility Analytics Institute analyst possessing extensive experience in customer-facing leadership roles in the utility and telecommunications industries.