Leaders and business intelligence at credit unions are putting atremendous focus on ways to use advanced data analytics to identifytrends, detect patterns and glean other valuable findings from thesea of information available to them. Without question, member datais valuable. But the greatest value lies in the ability to empowereach line of business to achieve strategic initiatives andperformance goals. When this empowerment is coupled with improvingmember service, a proven, repeatable best practice results.

Today's “analytics” are often visuals, graphs and trends whileadvanced analytics, while a journey, result inquantitatively-proven prescriptive statistics from regression andmachine learning. Advanced analytics enable the credit union toeffectively segment members to identify opportunities, improveservice and retention, and target new products and services.

Start the advanced analytics journey with the end goal in mindand consider these seven, “top-down” best practices as a roadmap toguide the credit union through each step.

1. Empower. Advanced analytics put actionableinformation in the hands of lenders and frontline employees. Thisdrives immediate, responsive actions that build member loyalty andhelp team members succeed. When the lines of business benefit frommeasured prescriptive action lists – management will view theselists as a critical tool to align the credit union's strategicgoals, create score cards, identify top performers and move thecorporate culture toward making more data driven decisions. Toprovide prescriptive action lists, the credit union should considerthe channel, message, timing and approach. The key to success withadvanced analytics is the ability to provide the right value forthe right member, at the right time, through the right channel,using the right message. Prescriptive statistics is beyond a simpleprobability based on past results. A prescriptive action list iscoveted by loan officers and branches because they have confidencethat taking action results in the best opportunity for success.

2. Refine: Once a set of successful strategiesand best practices for actions are set, it's time to further refinethose actions that lead to successful outcomes. This results in apredictable and positive outcome that improves member serviceand/or achieves growth, revenue, retention or risk managementgoals.

3. Apply: Is there a strategy to apply advancedanalytics? The opportunities to generate business value from dataand analytics are significant. Having a chief strategy officer andchief analytics officer on your team will allow you to curate yourmost critical information assets; an outsourced CSO/CAO may be agood option for many credit unions. Start by building statisticalmethods to measure the right action on the right segment, and usingthe right channel, message and timing. Measuring these resultswithin each segment is key. Segments should be based oncontribution and other algorithms coupled with usage determinationsthat assign the member to millennial, executive andmore. A feedback loop results by using these statistics asthe credit union further segments and refines the message, channel,timing and approach to create a set of recurring best practices.Random Forest, logistic and linear regression may reveal a portionof the segment that responds well to the action to then isolatethat portion of the segment and repeat, which results in futureprescriptive action lists.

4. Take action. Once key data segments havebeen identified, can the credit union take action using digitalchannels in addition to traditional branch, statement and printeddirect mail marketing? Digital channels include: SMS/text, email,mobile and internet banking calls for action. The credit unionshould be able to monitor if the member clicks and answers the“call to action” on a specific channel and sets as theirpreference. It's important to track, monitor and measure thesuccess of incremental actions and campaigns that result inrecurring best practices. Lines of business will be able to uncoverthe best actions, repeat and refine them over time for continuousimprovement and results.

5. Compute: Is there a process in place tocompute the data? How the credit union computes the necessaryresults to properly segment data will lead to improveddecision-making capabilities and positive business outcomes.Examples of these results include member contribution and how muchthe member is engaged. An extreme minority of members provide anextreme majority of contribution – typically the one to fivepercentiles provide 100% to 220% of total contribution. Percentilesfive to 89 often net to zero while the bottom 10% generate anegative contribution that leaves the 100%. Different actions areprudent within these different segments. The credit union needs tohave an effective way to compute advanced analytic algorithms fromthe data to identify trends and extract actionable information,enabling the credit union's drive to transformation.

6. Consume: How will the data be consumed?After the data is collected and cleansed, the credit union uses aBI platform, or reports, to present the information for consumptionby the end user. The information needs to deliver timelyinformation. Consumption of the data is critical in transformingthe member's data into an asset so lines of business can makedata-driven decisions, leading to improved member service andgrowth.

7. Lead: Having the right talent, team andtechnology makes all the difference. Many credit unions that planto, or already have implemented an analytics department have doneso by leveraging existing management and personnel. They soonrealize the hard way, they need to bring in fresh, specializedleadership. Start this process by identifying skill gaps,formalizing leadership roles and implementing a hiring, training orvetting process for outsourced expertise. Many credit unions thatpossess even the most advanced analytics capabilities cannot hireenough people to generate the right insights from member data. Besure the credit union has a knowledgeable data analytics team thatcan properly collect, prepare, manage and analyze the data frommultiple silos.

The biggest financial institutions have poured millions ofdollars into their advanced analytics investment by hiring teams ofanalysts and data scientists. They are building massive platformsto extract value from member/customer data. Since your best memberscould be the competition's best targets, it's important to usethese best practices in your advanced analytics roadmap. Make sureyou are ready to compete by making more data driven decisionsthrough analytics to find new ways to truly understand members'needs and optimize business results.

Steve D. Simpson, Ph.D. is CEO & ChiefData Scientist for FinTech Data Science + Consulting. He can bereached at 786-859-4100 or [email protected].

Paul Ablack is CEO & Founder ofOnApproach. He can be reached at 888-523-6121or [email protected].