Why do some data and analytics projects fail, while others go on to produce significant business outcomes? Today, many financial institutions are actively using data analytics to turn their data into actionable and profitable insights. However, the reality is most analytics projects do not always translate into easy success and big wins.  Too often, these projects will hit a plateau, unable to deliver the pot of gold or even a positive return on investment. So why the sub-optimal results?

Since deploying a data and analytics program is a complex process, there are many pitfalls and mistakes that you can make. While this list is by no means exhaustive, below are some of the most common mistakes we’ve seen as financial institutions embark on the data analytics journey.

Running before walking; walking before crawling:

Like most major projects, realistic goals and timelines need to be adhered to for the greatest chance of success.  I call this the crawl, walk, run phase of the data analytics project. The “crawl” phase is where the organization stops living and dying by their Excel VLOOKUPs and starts to use relational databases and common reporting tools.  During the “walk” phase of analytics is where the real analytics can begin. Your staff does not need to go to ten different places to get data for a report.  Near real-time analytics can be developed.  The “run” phase of analytics is where data science and statistical models become fully realized and leveraged.  During this time the data is working in every way imaginable and the organization has the structure, the culture, and the skills to make full use of the data’s potential.

Ignoring KPIs and success measures:

When you’re just getting started, it can be tempting to focus on small wins. However, it’s important to establish metrics like new business opportunities, customer satisfaction, onboarding, marketing, etc. Raw data must be turned into actionable information for it to have any real meaning.  That is why we emphasize the importance of establishing well-defined Key Performance Indicators (KPIs). KPIs are the quantifiable measures that a financial institution uses to gauge its strategic progress. The key is to keep the success measures simple, practical and relevant to the organization.  This is what will help you turn raw data into actionable, useful pieces of information so you can continually refine your KPIs for ongoing success.

Putting technology before strategy: 

When embarking on a data analytics project, one piece of advice is to not let IT drive the program. Many financial technology firms are offering best-of-breed software solutions that have features and functionality that are typically under-utilized or forgotten about in the urgency of deployment. In most cases 60-90% of the product features are often unused despite paying for 100% of the product. Clearly understand and outline the long-term strategic goals of the financial institution to identify and select technology solutions that fit your data analytics goals today and tomorrow. Otherwise, you run the risk of investing in software that doesn’t fit the financial institution’s vision, and the cost of converting later can be significant.

Overlooking data quality:

Data scientists know the importance of accurate and complete data. After all, if the data itself is unreliable, you’ll wind up making invalid conclusions based on your analysis. To avoid that pitfall, it is incumbent to spend the time and effort to diligently prepare and clean the data coming from all sources. This includes a broad range of cleansing such as: incorrect values, typos, aliases, inconsistencies, duplicate entries, outdated consumer information.  Your data need not be (nor will it ever be!) 100% clean to get started with your project.  It is, however, important to establish policies and procedures to identify and clean bad data on an ongoing basis.

Lacking data governance:

Think of data governance as a set of rules for inputting and maintaining data. It is a continuous quality control discipline that governs the overall management, usage, storage, monitoring, and protection of the financial institution’s data. Without dedicated governance processes, overtime, poor data quality will lead to inferior service delivery, reduced employee productivity, missed opportunities, and increased costs.  There are several subtopics in data governance that we have shared previously.  You can tackle data governance in a variety of different ways – just be sure not to overlook it in your analytics initiative.

Underutilizing talent:

Enlist the help of an experienced Chief Data Officer to keep everyone accountable and aligned with the overall strategic business objectives associated with your data and analytics projects.  Whether hiring a full-time executive or an outsourced expert like The Knowlton Group (who can resist a shameless self-plug!), view them as a trusted advisor and strategic partner.  Consultants and data scientists will work with your data analytics teams to pour through the data and train your staff to uncover and act on new opportunities that lead to your desired goals. These experts can advise decision-makers, monitor progress, track success and establish best practices that will position your data analytics project for success.

Underestimating the journey:

A fully-developed analytics program can take 18-to-36 months to deploy in its entirety. Deploying an advanced analytics initiative and building a data culture takes time, and many executives don’t plan for the long-term commitment.  Turning the data generated by consumers into actionable, comprehensive insights is a big task. The initial steps should include establishing short-term and long-term business goals. Once the strategy and business goals are in place, the technology infrastructure needs to have an effective and rigorous process to collect, cleanse, compute and consume the data. Staff needs to be assigned—and with the right skill-set to then use this complex data for discovery of insights and interpret results that lead to business opportunities.

What mistakes has your organization experienced while deploying your data and analytics program?

Next Steps

This list of pitfalls provides just a glimpse of the many mistakes financial institutions make when embarking on their data and analytics journey.   Before implementing any analytics solution, spend the necessary time determining how you would like data to strategically support the financial institution. Focus first on your KPIs and success measures and then explore your technological, operational, and deployment needs. With this discovery completed, you will have the foundation of what will be a working roadmap for your data and analytics journey.

Need expert help planning, designing, and implementing your data and analytics strategies and solutions? Contact The Knowlton Group today by email at [email protected] to learn more!

If you have reviewed the important first six steps outlined in our first article on Building an Analytics Roadmap, then you are ready for the final steps to complete the journey.

To recap, we firmly believe that analytics in the bank and credit union industry will become necessary for survival.  Those who adopt early will see significant competitive advantages.  However, those who adopt without careful planning, strategy, execution and commitment will see their time and money wasted due to poor data governance, inadequate technology, lack of data talent, and uncommitted or misaligned leadership.   These final steps in building a solid roadmap will get you on your way to reaching your BI goals.

Step 6: Assess & upgrade technology capabilities:

Nothing can derail your data analytics project than a lack of the right hardware, software and data analytics technology. Create an assessment of your technology inventory as part of the roadmap and indicate the need for further technology capabilities that align with your overall strategy.  Assess if complex infrastructure and security is already in place or if the analytics systems require significant infrastructure updates including robust security, abundant data storage, and other components of a solid infrastructure.

Step 7: Ensuring data quality:

Data quality can vary significantly depending on how it was collected, stored, cleansed, and processed. Establishing data quality checks ensures that data is complete, timely, and accurate. Implementing a comprehensive data profiling effort and data quality initiative can help identify data quality issues earlier in the implementation timeline.  Working with subject matter experts from the various business units can help identify and clean up any issues that arise.

Step 8:  Identify analytics success measures:

Many organizations are excited to get started building out a data warehouse and establishing an analytics program.  However, most fail, at the outset, to establish measures that define success (or failure) for the analytics initiative.  These can be ROI-related or measures around adoption and utilization of analytics throughout the organization.

Establishing these success measures at the start of process allows management to gauge whether or not the program is meeting expectations.  If success is not being met, management and the BI team can take corrective actions to re-align the analytics program with expectations.

Step 9: Create an analytics workplace culture:

Having an analytics-ready culture is a huge indicator of future success. FIs that have taken the time to create a workplace culture that understands, visualizes, and believes in the capabilities of the data will see results. Establishing a culture in which everyone — including your leadership — understands and appreciates how data brings value to business initiatives will ensure that your processes and culture are aligned around producing profitable insights and positive business outcomes.  Getting the best value from your data means you must first trust it yourself and then give employees the opportunity to learn and embrace data-driven decision making. (2)

 Step 10: Create a data dictionary:

A data dictionary provides detailed information about each data element within your analytics environment.  Users need to know what the data fields and metrics mean. When you have a clear list of metrics and their definitions, it helps to eliminate assumptions, hours of guesswork, errors, and confusion.  Take the time to generate the data dictionary with clear, unambiguous and agreed-upon definitions. This requires collaboration of all the key players within the data analytics project.

Step 11:  Provide ongoing training:

You would never implement a new LOS without training your staff on how to use it, right?  Similarly, an organization won’t have success with analytics without providing proper and consistent training.  Business users need to be properly trained on the data elements included in the data model, how to access information, and the various other aspects of the analytics program.

Step 12:  Prioritize implementation phases:

A successful data analytics roadmap should divide the total implementation into logical phases to help reduce the risk of failure.  Start with the highest priority initiative(s) and refine the initial scope to mitigate the downside risks. Set goals and milestones by breaking up the entire project into manageable phases. This approach not only helps manage development of the solution, but it also gives the analytics team time to train business users on how to leverage each subsequent phase of the implementation.  Failing to do this is one of the most common reasons that BI projects fail.

Are you ready to start building your FI’s data analytics roadmap?  The twelve steps covered in the past two articles may seem daunting.  The Knowlton Group can help.   Our expertise, years of deep experience working with FIs on assessing and implementing proven data analytics strategies and solution can work for you.   Contact us today to learn how we can help assist your BI strategy or assess your current goals to make sure your capabilities are aligned.

Resources
2. 4 Strategies to Create a Data-Driven Company Culture