CU Employee Community Chair

Explore Vizualization - For Credit Unions


The Webinar on Explore Visualization- Taking Game to the Next Level for Credit Unions, was conducted by Karan Bhalla and Austin Wentzlaff. The basics include the volume of the data, the information created every second with this data, the value of big data and better analytical insights. The different steps and levels of data analytics start with Data Access then moving on to Data Management, Reporting, Forecasting and finally Predictive Modeling and Optimization. Almost all credit unions are in the first two levels. The difference between Data Analytics and Visualization can be recognized as where data analytics is about recognizing patterns and deriving meaning from complex data sets while identifying underlying models and patterns that feed visualization, visualization is about presenting patterns graphically for stakeholders to understand difficult concepts while identifying areas that need improvement and attention to feed decision making and discovery process. While Data Analytics is descriptive, predictive and prescriptive supported by tools such as SPSS, SAS, R and Python; visualization is interactive and static supported by tools such as SSRS, Tableau, PowerBI, Qlikview, and Domo.

The current state of data is unorganized and needs to be cleaned up. There is a lot of data at hand and some of the examples include Membership Data, Transaction and ACH Data, Bureau Data, Online/External Data, and Product Data. The test is how should this data be used, how should it be executed and how do we include member experience. The challenge with the existing mechanism is that it involves labor-intensive manual process among other challenges of redundancy, data existing throughout different separate data silos, visual impact, accessibility, and accuracy because human intervention leads to human error. The proposed architecture for data organizing strategies is the M360 Data Model where data is taken from different ancillary systems, organized in M360 data model and put into the standardized format. It is important what visualization tools are being used and what they are being used for, the predictive models being tapped into and other applications or data services referred to as the process of getting it. The CUs should keep in mind some checks and balances to allow for it.

There are mainly three visualization tools which are most commonly invested into, they being SSRS, Tableau, and PowerBI. SSRS is Sequel Server Reporting Services, and though it is not the most attractive tool, it involves no cost and skills. Tableau can be used for large data sets with basic statistical tools. It is flexible and attractive and well integrated with R. PowerBI is inexpensive and integrated with Azure and SQL server and even R. The other tools include Qlikview and Domo (cloud-based and catchy). Some custom build tools are Java, R,, and Python.

There are four requisites to create effective visualizations:

  1. Make it Relatable which means delivering right messages to the right stakeholders.
  2. Quicker the Better is providing relevant information in few seconds.
  3. Looks Matter is giving key information on top. It is about design thinking and presentation.
  4. Less is More means to not add unnecessary data.

Some fundamental visualization FAQs consist of where to start which is assessing the current state as to what are the true data needs, setting goals and objectives as to getting the right information to the right users and establishing requirements as to what you hope to achieve, technical and business requirements. How to decide on a tool should start with existing data and technology assessment, compatibility check with the business objective and resource planning. How to get maximum from the tool can be obtained by following the 3S’s, promoting data-driven thinking and ongoing review of ROI. The result of the process of adopting and investing in visualization is being organized.

The 3S’s of Visualization is Standardization, Simplicity, and Scalability. Embracing standardization is creating and accessing the same data across the model whilst reducing complexities and disparities. It means gathering data from different sources, creating single truth, standard dashboard and providing easy user access. Fueling simplicity is interpretable, easy to use and customizable data that facilitates decision making. Applying scalability is providing an enterprise-wide solution, efficient data model, accommodating large volume and processing speed. It means gathering data from different sources, standard data structure, standard dashboards and ultimately adding more users.

The ORBIT Approach should be adopted while going for visualization, which is, Organize data, Review and pick the right tools, Build simple scalable and standard reports and dashboards, Identify areas for improvement and attention and most importantly Take action and do not sit on the results.

There are few questions which are relevant in this regard. Like how big should the CU be to use visualization tools like the number of members, asset size, and the number of branches and can a small CU be also benefited by it? The answer to this is that visualization is for all credit unions and it is just a matter of getting a start; an example can be 700-400,000 members and 40 million to 4.7 billion in assets. Secondly, how similar or different are these visualization tools from data warehouse tools. The two are entirely different, where data warehousing tools are technologies used to organize and store, and requires a different skill set; visualization is viewing data and using it and making it more compelling. Thirdly, what kind and how big of a team is required at the CU to support visualization? It depends on the tools used, and no huge team is required for it. It depends on how efficient and successful of a data management solution is put into place. If that’s in place, then the team can be very small as well. Lastly, what is an ideal budget for visualization and how long would it take to have a sustainable BI process in place? It depends on the amount effort and works put into it. Early reporting stage is done at this point. It is difficult to find good visualization folks and the key is to maximize what you can get your partners and be strategic about planning that resource. It is all about what your approach is going to be, will it be self-made or outsourced through partners, is what is going to make the impact.


E-mail me when people leave their comments –

You need to be a member of Credit Union Data Analytics and Digital Transformation Leader to add comments!

Join Credit Union Data Analytics and Digital Transformation Leader



Related Post