top of page
Search

The Essential Elements of Analytics Success

Updated: Jul 22, 2021

“Data is the new oil.”  That’s a now-famous quote from Clive Humbly, a British mathematician and data scientist. You’ve probably heard it before and agreed.  Everything today, seemingly, is powered by data. As a bank executive, you are likely overwhelmed with inquiries from endless vendors that want to help you unlock the knowledge your data possesses with their form of “magic.” 

We’ve heard all the hype too and in this article, we want to help bring clarity to the overall purpose and uses of analytics.  Over the course of a few articles, I want to convey what I’ve learned in my thirty-five-year career of developing and deploying successful analytics solutions to help you get started on the right foot.

  

Let’s start with defining “analytics success.”  Many companies fall into a trap when starting on their analytics journey, assuming that developing dashboards and models from previously underutilized data is the objective.   Now granted, creating dashboards, like the one shown here (designed by Tableau and their partner Slalom) that summarizes the current performance of a loan portfolio, is an important goal. Still, it’s only a step towards success. Dashboards, or data products, are tools, not answers.



Analytics success is only achieved when these tools lead to significant reductions in the time it takes to make decisions, the quality of those decisions, and a growing number of people using them to help make decisions.   


The Analytics Progression


There are many ways in which companies use analytics to improve decision-making.  Assessing where you are today is necessary to plan your path forward.  What’s required of the enterprise to perform different analytics functions varies greatly.  Creating a roadmap to the future needs to account for developing the capabilities needed at each stage. 

In their seminal book, Competing on Analytics, Tom Davenport and Jeanne Harris described a series of analytics types that progress from essential, yet straightforward, to sophisticated and powerful: 

  • Descriptive Analytics is the world of reporting.  It’s a fundamental part of any bank’s operations but limiting.  Static reporting is akin to driving down the road while looking in the rear-view mirror. Reports provide insight but can create a false sense of security.  Knowing how many loans in your portfolio are delinquent is helpful but doesn’t help you understand what to do about them. 

  • Diagnostic Analytics is reporting on steroids.  It enables analysts to dig into the detail behind the summarized numbers.  Loan delinquencies are up, but where? What types of loans? Are specific industries driving the change?  The dashboard shown above isn’t limited to conveying only a summarized view of this data.  You can click on many of the fields to display the detail behind the numbers. All in one place and with one click.  That’s powerful! Once attributes are evaluated, you can form credible hypotheses about what’s causing the problems and acting to correct them.  Moving from a report-dominated culture to one driven by diagnostics is a giant leap forward for many banks.  The good news is that your bank is probably already practicing some form of diagnostics. But there is perhaps much to gain by improving the speed at which data is delivered and its granularity.  


  • Predictive Analytics is a significant step forward from diagnostics and require a much different set of skills and capabilities. Predictive analytics utilizes historical data and the relationships within it to model what is likely to happen given today’s inputs of the model variables.  Predicting the probability that a borrower will default on a loan is the most common use of predictive analytics in banking.  Developing predictive models requires the expertise of statisticians, mathematicians, or data scientists.  It also places an increasing burden on making abundant, high-quality data available to them.

  

  • Prescriptive Analytics “kicks it up a notch.” It allows us to evaluate outcomes across a range of future conditions and constraints to help us make the optimal decision today, considering many uncertainties.  Prescriptive analytics answers questions like: “What should the mix of industries in our loan portfolio be?”  Or ”Given today’s mix of exposures, how likely are we to need more capital under different economic scenarios?”.  

Going back to the analytics progression chart above, Davenport and Harris contend that mastering how to create value from increasingly sophisticated analytics types is where you can establish competitive advantages for your bank.  I agree wholeheartedly.  Additionally, I contend that analytics success is no longer restricted to large banks with seemingly infinite resources.

While big banks may have people and money to throw at problems, their data environments are often highly complex, and coordinating all the functional siloes required to move forward can be daunting.  These are just two of the reasons that financial institutions of all sorts report the existence of an “analytics gap.”  For many, if not most banks, insurance companies, investment banks, and the like, the gap between the promise of what analytics can deliver and what is taken to the bottom line is significant. 


Data may well be the new oil, but as Michael Palmer added: “Data is valuable, but if unrefined, it can’t be used.”  In and of itself, data is raw material to analytics. 


Future Articles on Analytics


In future articles, I will describe what I, and many others, have found to be the keys to realizing as much value as possible from your analytics investments.  Here’s a list of what’s coming: 

  • Decisions – What decisions are you trying to improve with data? 

  • Data – Is your data accessible, of breadth and depth, and high-quality? 

  • Talent – What skills are needed? How are they established and improved? 

  • Operations – What tools, processes, and infrastructure are required to enable data-driven decisions? 

  • Culture – How do you create a data-driven culture across your enterprise? 

  • Roadmap – How do I integrate all of what needs to be developed into a coherent and achievable plan? 

About FinResults


At FinResults, we have the methods, experience, and tools to help you develop an analytics roadmap and implement it as your successful analytics strategy.  With a Strategic Analytics Assessment as the first step, we can tell you what is needed in the domains listed above to achieve analytics success.   


FinResults is an innovation partner for banks, founded by bankers with years of digital and data management implementation experience. We can bring not only experience but also the technical resources, agile implementation, experience with digital transformation, and advanced data environments to help your bank be successful with its transformation initiatives. 


For Further Information


For more information, please contact us at info@finresults.com or +1 (561) 288-6548.



About the Author:


Tom Warden is an experienced leader of strategy and innovation across various complex business problems solved with data, analytics, artificial intelligence, and thought leadership. As a trusted advisor to C-suite executives in mid-sized and Fortune 100 companies, Tom develops plans and builds teams to improve the quality and efficiency of how data is used enterprise-wide across varying functions to make critical business decisions. Tom creates game-changing competitive advantages through emerging technologies to boost revenue and capture market opportunities. Tom earned an MBA from Harvard University and a Bachelor of Science from The Ohio State University, where he majored in Accounting. Tom can be reached at tom.warden@finresults.com.





Comments


bottom of page