data - CULytics Community2024-03-28T14:42:34Zhttps://culytics.com/blogs/feed/tag/dataMastering Time Period Bias: Enhancing Data Accuracy in Credit Union Decision-Makinghttps://culytics.com/blogs/time-period-bias-in-credit-union2023-12-06T16:51:59.000Z2023-12-06T16:51:59.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"> </p>
<p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12309968461,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12309968461,RESIZE_710x{{/staticFileLink}}" width="710" alt="12309968461?profile=RESIZE_710x" /></a></p>
<p class="zw-paragraph heading0">Time Period bias, also known as period bias, refers to the distortions that arise when data or outcomes are affected by the specific time or period in which they're observed or collected. For credit unions, understanding and accounting for time period bias is crucial for accurate decision-making and member service. Here are some examples of time period bias in the credit union context:</p>
<ol>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Seasonal Loan Applications</strong>: If a credit union evaluates the success of a new loan product based solely on applications received during a holiday season, they might mistakenly believe the product is always in high demand, overlooking the seasonal influence on borrowing.</p>
</li>
</ol>
<ol start="2">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Interest Rate Changes</strong>: Assessing the popularity of savings accounts after a recent interest rate increase might lead to an overly optimistic view of the product's general attractiveness if not accounting for the rate change's temporary influence.</p>
</li>
</ol>
<ol start="3">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Post-Crisis Analysis</strong>: Evaluating member financial behaviors or loan default rates immediately after an economic downturn without considering the broader economic context can lead to misinterpretations of member reliability.</p>
</li>
</ol>
<ol start="4">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Tax Season Fluctuations</strong>: Observing a spike in account deposits during tax refund season and assuming this reflects a general trend in member savings behavior can be misleading.</p>
</li>
</ol>
<ol start="5">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Promotional Periods</strong>: If a credit union offers special promotions or bonuses for opening new accounts and evaluates the product's success solely during this promotional period, they might have a skewed perception of the product's long-term appeal.</p>
</li>
</ol>
<ol start="6">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Post-Event Surveys</strong>: Surveying members immediately after a positive community event sponsored by the credit union might yield more favorable general feedback about the credit union due to the recency of the positive experience.</p>
</li>
</ol>
<ol start="7">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Digital Platform Uptake</strong>: Launching a new digital platform or app feature around the same time when there's an increased need for online banking (e.g., during a pandemic) might lead to an overestimation of its inherent popularity.</p>
</li>
</ol>
<ol start="8">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Branch Traffic Post-Opening</strong>: Evaluating the traffic and success of a new branch in its initial months without considering the novelty effect can lead to incorrect assumptions about its long-term viability.</p>
</li>
</ol>
<ol start="9">
<li>
<p class="zw-list zw-paragraph heading0"><strong>End-of-Year Financial Products:</strong> Some members might rush to open retirement accounts or make specific financial moves at the end of the tax year. Relying on data from this period alone can distort the understanding of product demand throughout the year.</p>
</li>
</ol>
<ol start="10">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Member Feedback During System Upgrades:</strong> If a credit union gathers feedback during a period of significant system upgrades or maintenance, the feedback might be disproportionately negative due to the temporary inconvenience, not reflecting general member satisfaction.</p>
</li>
</ol>
<p>To address time period bias, credit unions should aim to collect data over extended periods, encompassing various conditions and events. They should also continually assess the broader context in which data is gathered and be wary of making broad generalizations from short-term observations.</p></div>Elevate Your Credit Union with Data Analytics Expertisehttps://culytics.com/blogs/elevate-your-cu-with-data-analytics-expertise2023-11-21T19:31:46.000Z2023-11-21T19:31:46.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12296647083,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12296647083,RESIZE_710x{{/staticFileLink}}" width="710" alt="12296647083?profile=RESIZE_710x" /></a>At CULytics, we recognize the game-changing potential of data analytics for credit unions. It has the power to enhance member services, streamline operations, and fuel growth. Yet, a significant challenge that many credit unions face during data analytics implementation is "Inadequate Education."</p>
<p class="zw-paragraph heading0">Bridging this educational gap is essential to ensure that employees across all organizational levels grasp the value of data analytics and wield its insights effectively. Here are some tips about how you can overcome this challenge and harness the full power of data analytics.</p>
<ol>
<li class="zw-paragraph heading0"><strong>Training and Workshops<br /></strong>Your journey begins with comprehensive training sessions and workshops, laying a strong foundation for data analytics skills within your team. Training should cover a spectrum of topics, including industry success stories, data analysis tools, data visualization techniques, and the interpretation of analytical results. Include hands-on training with real data scenarios, enabling employees to apply their learning to practical situations seamlessly.<br /><br /></li>
<li class="zw-paragraph heading0"><strong>Data Literacy Programs<br /></strong>Data literacy is the bedrock upon which effective data analytics rests. Develop data literacy programs, focusing on enhancing understanding of data concepts and terminology. Ensure that all your team members are comfortable with key data-related terms like "data analytics," "data visualization," "KPIs," and "predictive modeling." Furthermore, provide a glossary of these terms, making data literacy inclusive and easily accessible to all members of your team.<br /><br /></li>
<li class="zw-paragraph heading0"><strong>Communication of Success Stories<br /></strong>Real-world success stories and case studies are potent tools to showcase the tangible impact of data analytics. Demonstrate how analytics has led to better decision-making and improved member services. Additionally, highlight examples of how data analytics has empowered credit unions, like yours, to identify opportunities and successfully overcome specific challenges.<br /><br /></li>
<li class="zw-paragraph heading0"><strong>Leadership Involvement<br /></strong>Leadership support is crucial for the success of data analytics initiatives. When top leadership actively champions and participates in data analytics projects, it sets a motivating example for others to follow. Ensure that leadership effectively communicates the vital importance of data analytics in aligning with your credit union's mission and strategic goals.<br /><br /></li>
<li class="zw-paragraph heading0"><strong>Regular Updates and Reporting<br /></strong>Establish robust reporting mechanisms to keep your employees informed about the progress and outcomes of data analytics projects. Share key performance indicators (KPIs) and metrics that vividly convey the value generated through data analytics.</li>
</ol>
<p class="zw-paragraph heading0">By implementing these strategies, your credit unions can cultivate a culture that reveres data analytics, fosters profound understanding, and catalyzes the effective utilization of data-driven insights across the organization.</p>
<p>Unlock the full potential of data analytics for your credit union with CULytics. Contact us today at <strong>info@culytics.com</strong> to learn more about how we can assist you in bridging the communication and education gap, and embarking on a journey toward data-driven excellence.</p></div>Unmasking Reporting Bias: Navigating the Credit Union Landscapehttps://culytics.com/blogs/unmasking-reporting-bias2023-11-14T18:29:55.000Z2023-11-14T18:29:55.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><strong><a href="{{#staticFileLink}}12291613462,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12291613462,RESIZE_710x{{/staticFileLink}}" width="710" alt="12291613462?profile=RESIZE_710x" /></a></strong></p>
<p class="zw-paragraph heading0"><strong>Reporting Bias</strong> occurs when the frequency or manner in which data is reported or recorded is influenced by a range of factors, leading to an incomplete or skewed representation of information. Within the context of a credit union, reporting bias can be particularly impactful, affecting decisions and insights related to member needs, risk assessment, and overall strategy. Here are some examples of reporting bias in the credit union space:</p>
<ol>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Member Feedback</strong>: If a credit union only actively solicits and publicizes positive feedback or testimonials about its services, while not addressing or acknowledging negative experiences, it can give a skewed impression of member satisfaction.</p>
</li>
</ol>
<ol start="2">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Loan Default Rates</strong>: A credit union might underreport the number of loan defaults or delinquencies to project a healthier financial image, which would mislead stakeholders about the institution's actual risk profile.</p>
</li>
</ol>
<ol start="3">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Online Service Glitches</strong>: Suppose members frequently encounter technical glitches when using the credit union's online services. If only a few report these issues, the credit union might underestimate the problem's magnitude and not prioritize fixing it.</p>
</li>
</ol>
<ol start="4">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Financial Education</strong> Outcomes: After hosting financial education seminars, a credit union might primarily highlight success stories or cases where members improved their financial behavior, neglecting instances where no noticeable change occurred.</p>
</li>
</ol>
<ol start="5">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Survey Participation</strong>: Members who have strong feelings about the credit union, either positive or negative, might be more likely to respond to satisfaction surveys. This self-selection can lead to an overrepresentation of extreme opinions and an underrepresentation of moderate or indifferent views.</p>
</li>
</ol>
<ol start="6">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Credit Card Usage</strong>: If a credit union offers rewards for certain credit card transactions, members might be more likely to report or highlight these specific transactions, thereby giving an incomplete picture of overall card usage.</p>
</li>
</ol>
<ol start="7">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Investment Returns</strong>: A credit union might emphasize high-performing investments in its communications to members, downplaying or omitting underperforming assets, which can mislead members regarding the overall portfolio's performance.</p>
</li>
</ol>
<ol start="8">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Security Incidents</strong>: A credit union might not disclose or underreport minor security breaches or incidents to maintain member trust, even if these incidents highlight vulnerabilities that need addressing.</p>
</li>
</ol>
<ol start="9">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Branch Usage Data</strong>: If a branch in a prime location has high foot traffic but most interactions are basic transactions (e.g., cash withdrawals), while another branch has fewer visitors but more high-value interactions (e.g., loan consultations), only reporting raw foot traffic numbers can give a skewed picture of branch value.</p>
</li>
</ol>
<ol start="10">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Member Demographics</strong>: In its annual report, a credit union might highlight the diversity of its membership based on a few select criteria, ignoring other aspects of diversity, which can lead to an incomplete understanding of its member base.</p>
</li>
</ol>
<p>Addressing reporting bias requires a commitment to transparency, robust data collection methods, and periodic audits or reviews to ensure that data representation is accurate and complete. For credit unions, overcoming this bias is essential to maintain trust with members and make informed decisions.</p></div>Unleash the Power of Real-Time Data: Top Use Cases for Credit Union Leadershttps://culytics.com/blogs/unleash-the-power-of-real-time-data-use-cases2023-11-01T14:36:19.000Z2023-11-01T14:36:19.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12281594679,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12281594679,RESIZE_710x{{/staticFileLink}}" width="710" alt="12281594679?profile=RESIZE_710x" /></a></p>
<p class="zw-paragraph heading0"> </p>
<p class="zw-paragraph heading0">In a recent data analytics roundtable session, we delved deep into the realm of real-time data analysis and its transformative potential for credit unions. The discussion brought to light several compelling use cases where real-time data can truly make a difference. Today, we're excited to share these insights with you, as they have the potential to revolutionize the way credit unions serve their members and prospects.</p>
<p class="zw-paragraph heading0"><strong>1. Real-Time Help for Seamless Member Onboarding</strong></p>
<p class="zw-paragraph heading0">Imagine a scenario where a member or prospect is in the midst of applying for a product or service, and they encounter an issue or have a question. Real-time data analytics can provide immediate assistance, reducing application abandonment rates. By offering timely support, credit unions can ensure a smooth and frustration-free onboarding process, enhancing member satisfaction.</p>
<p class="zw-paragraph heading0"><strong>2. Enhancing Security with Real-Time Risk and Fraud Detection</strong></p>
<p class="zw-paragraph heading0">Security is paramount in the financial sector, and real-time data analysis is a game-changer in this regard. By continuously monitoring transactions and member activities, credit unions can swiftly detect and respond to suspicious behavior. This proactive approach not only safeguards your organization but also reassures members that their financial well-being is a top priority.</p>
<p class="zw-paragraph heading0"><strong>3. Harnessing Real-Time Feedback Collection</strong></p>
<p class="zw-paragraph heading0">In the age of instant gratification, real-time feedback collection is a powerful tool for credit unions. Members and prospects are more likely to provide feedback when it's convenient for them, leading to more accurate insights. By actively seeking and acting upon real-time feedback, credit unions can fine-tune their offerings, improving member experiences and loyalty.</p>
<p class="zw-paragraph heading0"><strong>4. Empowering Members with Real-Time Application Status Updates</strong></p>
<p class="zw-paragraph heading0">The waiting game can be frustrating for members who have applied for loans, credit cards, or other financial products. Real-time data access allows members to check the status of their applications instantly. This transparency builds trust and provides peace of mind, demonstrating that credit unions value their members' time and are committed to keeping them informed.</p>
<p class="zw-paragraph heading0"><strong>5. Personalized Marketing and Member Engagement</strong></p>
<p class="zw-paragraph heading0">Real-time data analysis can help credit unions understand their members' preferences and behaviors better. By tracking member interactions across various channels and touchpoints, credit unions can create personalized marketing campaigns and offers that resonate with individual members. This not only increases the likelihood of cross-selling and upselling but also strengthens member loyalty.</p>
<p class="zw-paragraph heading0"><strong>6. Real-Time Loan Approval and Decisioning</strong></p>
<p class="zw-paragraph heading0">Members often require quick loan approvals for various needs. Real-time data analysis enables credit unions to assess member eligibility and risk instantly. This agility in decision-making not only enhances member satisfaction but also allows credit unions to capitalize on lending opportunities promptly.</p>
<p class="zw-paragraph heading0"><strong>7. Dynamic Pricing and Rate Adjustments</strong></p>
<p class="zw-paragraph heading0">In a constantly changing financial landscape, it's essential to have a competitive edge. Real-time data can help credit unions adjust their pricing and interest rates based on market conditions, member profiles, and risk assessments. This flexibility ensures that credit unions remain competitive while optimizing their profitability.</p>
<p class="zw-paragraph heading0">These seven use cases underscore the potential of real-time data analysis to drive member-centricity, efficiency, security, and competitiveness within credit unions. Embracing this technology can set your organization apart in a competitive landscape, delivering superior service and peace of mind to your members.</p>
<p>In the coming weeks, we will explore each of these use cases in greater detail, providing actionable insights and best practices to help you harness the full potential of real-time data.Till then feel free to register for the upcoming roundtable here - <strong><a href="https://events.culytics.com/da-nov" target="_blank">https://events.culytics.com/da-nov</a></strong></p></div>Navigating Missing Data in Credit Unions: Methods, Biases, and Mitigation Strategieshttps://culytics.com/blogs/navigating-missing-data-in-credit-unions2023-10-11T18:42:22.000Z2023-10-11T18:42:22.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12253841280,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12253841280,RESIZE_710x{{/staticFileLink}}" width="710" alt="12253841280?profile=RESIZE_710x" /></a></p>
<p class="zw-paragraph heading0">Handling missing data is a significant challenge in many industries, including in the financial and credit union sectors. Incorrect or biased treatment of missing data can have severe consequences, such as unfair loan approvals or skewed financial risk assessments.</p>
<p class="zw-paragraph heading0">Here are some methods used in the credit union world to handle missing data, along with the potential biases they might introduce:</p>
<ol>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Deletion (Listwise or Pairwise)</strong>:</p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> This involves removing entire observations (rows) where any single value is missing.</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> If the missing data isn't random, this can lead to a non-representative sample, especially if a particular group is more likely to have missing data. For instance, if young loan applicants tend not to have a credit history and get excluded, the resulting analysis will skew towards older individuals.</p>
</li>
</ul></li>
</ul>
<ol start="2">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Mean/Median/Mode Imputation:</strong></p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> Replace missing values with the mean (for continuous data), median (when data has outliers), or mode (for categorical data).</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> It assumes that the missing value is close to the average, which might not always be the case. This can underestimate variability and artificially inflate the number of data points at the mean/median/mode. For example, assuming missing incomes as the mean income might under-represent both low and high earners.</p>
</li>
</ul></li>
</ul>
<ol start="3">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Regression Imputation:</strong></p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> Use regression models to predict and impute missing values based on other related variables.</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> This assumes that there's a linear relationship between the missing data and other variables. If incorrect, this can lead to biased estimates.</p>
</li>
</ul></li>
</ul>
<ol start="4">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Last Observation Carried Forward (LOCF):</strong></p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> Used mainly in time-series data, where the last known value is used to fill subsequent missing values.</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> It assumes trends don't change, which can be problematic if there are fluctuations or shifts in the data over time. This can lead to incorrect trend analysis in financial time-series data.</p>
</li>
</ul></li>
</ul>
<ol start="5">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Stochastic Regression Imputation:</strong></p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> Similar to regression imputation, but includes a random residual term to account for prediction errors.</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> It's still based on the assumption that other variables can predict the missing value, and introducing randomness can sometimes add noise.</p>
</li>
</ul></li>
</ul>
<ol start="6">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Multiple Imputation:</strong></p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> Multiple datasets are created with different imputations, and analyses are performed on each to get combined results.</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> While this method reduces the uncertainty of imputed values, the choice of imputation model can still introduce bias if the underlying assumptions are incorrect.</p>
</li>
</ul></li>
</ul>
<ol start="7">
<li>
<p class="zw-list zw-paragraph heading0"><strong>K-Nearest Neighbors (KNN) Imputation:</strong></p>
</li>
</ol>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Description:</strong> Missing values are imputed using values from "k" similar observations.</p>
</li>
</ul></li>
</ul>
<ul>
<li><ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Potential Bias:</strong> If the choice of "k" or the distance metric isn't appropriate, it might lead to biased imputations.</p>
</li>
</ul></li>
</ul>
<p class="zw-paragraph heading0">In the credit union world, understanding the nature of missing data is crucial. For instance, if a person doesn't provide their monthly expenditure, is it because it's too high, too low, or they just forgot? The reason can greatly affect the method of imputation.</p>
<p class="zw-paragraph heading0">Lastly, biases can be mitigated by:</p>
<ul>
<li>
<p class="zw-list zw-paragraph heading0">Using advanced models like random forests or deep learning for imputation, which can capture complex patterns.</p>
</li>
</ul>
<ul>
<li>
<p class="zw-list zw-paragraph heading0">Regularly validating and cross-validating the data imputation methods against actual data to gauge their accuracy.</p>
</li>
<li>Maintaining transparency and open communication about the methods used so stakeholders understand potential risks and uncertainties.</li>
</ul></div>Uncovering Biases in Data Preprocessing: Insights from the Credit Union Contexthttps://culytics.com/blogs/uncovering-biases-in-data-preprocessing2023-10-04T16:05:56.000Z2023-10-04T16:05:56.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12239457660,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12239457660,RESIZE_710x{{/staticFileLink}}" width="710" alt="12239457660?profile=RESIZE_710x" /></a></p>
<p class="zw-paragraph heading0">Data preprocessing is a crucial step in the data analysis pipeline where raw data is cleaned and transformed to prepare it for analysis or modeling. While preprocessing can help in dealing with missing, inconsistent, or noisy data, it can also introduce biases if not done carefully. Here are some examples of data preprocessing bias, some of which can be illustrated using the credit union context:</p>
<ol>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Imputation of Missing Data</strong>: If a credit union imputes missing data on loan applications—say, filling in missing income values with the average income—it might introduce bias if the missingness isn't random. For example, if higher-income individuals are more likely to leave that field blank, the imputation could underestimate their income.</p>
</li>
</ol>
<ol start="2">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Feature Scaling</strong>: Standardizing or normalizing features (like income or loan amounts) can inadvertently give more weight to certain features over others in some algorithms, affecting the outcome of analyses or predictions.</p>
</li>
</ol>
<ol start="3">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Data Smoothing</strong>: While smoothing can help reduce noise in data, over-smoothing might eliminate genuine fluctuations or trends. For instance, smoothing out fluctuations in monthly deposits might miss genuine patterns, like seasonal effects.</p>
</li>
</ol>
<ol start="4">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Binning Continuous Variables</strong>: Converting a continuous variable, like age, into bins (e.g., 18-25, 26-35) can lead to loss of information and might introduce arbitrary boundaries. Two members aged 25 and 26 would be placed in separate bins, even though they're close in age.</p>
</li>
</ol>
<ol start="5">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Oversampling and Undersampling</strong>: To address class imbalance, like in a dataset where loan defaults are rare, one might oversample the default cases or undersample the non-default cases. While this can help models perform better, it can also introduce bias and affect the generalizability of the model.</p>
</li>
</ol>
<ol start="6">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Removing Outliers</strong>: If a credit union decides to remove all loan applications that request unusually high amounts, considering them outliers, it might inadvertently exclude genuine cases or specific segments of the population.</p>
</li>
</ol>
<ol start="7">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Feature Selection</strong>: Choosing which variables to include in a model based on some criteria might leave out important variables. If a credit union uses only employment status and income to predict loan default and ignores credit history, the model might be biased.</p>
</li>
</ol>
<ol start="8">
<li>
<p class="zw-list zw-paragraph heading0"><strong>One-Hot Encoding</strong>: When categorical variables are converted into binary columns, the increase in dimensionality can affect some models. If not handled correctly, this can lead to multicollinearity or overfitting.</p>
</li>
</ol>
<ol start="9">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Temporal Splitting</strong>: In time-series data, splitting data randomly for training and testing can lead to future information leaking into the past. For credit unions, this could mean using future financial data to predict past events, which is not realistic.</p>
</li>
</ol>
<ol start="10">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Ignoring Data Dependencies</strong>: If a credit union has multiple accounts for a single member and treats each account as an independent data point, it might ignore the inherent correlations between a member's accounts, leading to biased models.</p>
</li>
</ol>
<p>To mitigate data preprocessing biases, it's essential to understand the data, the context, and the implications of preprocessing decisions. Validating models or analyses on diverse datasets and continually re-evaluating preprocessing choices are also good practices.</p></div>Unveiling the Hidden Impact of Confirmation Bias in Credit Unionshttps://culytics.com/blogs/hidden-impact-of-confirmation-bias-in-credit-unions2023-09-19T14:33:47.000Z2023-09-19T14:33:47.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12227907860,RESIZE_584x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12227907860,RESIZE_584x{{/staticFileLink}}" alt="12227907860?profile=RESIZE_584x" width="568" /></a></p>
<p class="zw-paragraph heading0">In the dynamic landscape of credit unions, data-driven decision-making is essential for success. However, lurking beneath the surface of analytical insights lies a cognitive bias that can significantly impact outcomes: confirmation bias. This psychological phenomenon, where individuals tend to favor information that aligns with their preexisting beliefs, can cast a shadow on the accuracy and effectiveness of credit union strategies. In this article, we explore the realm of confirmation bias within credit unions, sharing real-life examples and shedding light on its potential pitfalls.</p>
<h2><strong>Real-Life Industry Cases:</strong></h2>
<p class="zw-paragraph heading0"><strong>Success Story:</strong> In the banking realm, Capital One embraced data analytics to optimize its credit card offerings. The bank adopted a comprehensive approach by analyzing diverse customer data points and behaviors. This included not only spending habits but also customer service interactions. By avoiding confirmation bias and seeking a holistic view, Capital One managed to successfully identify lucrative customer segments and design tailored credit card products that resonated with individual needs.</p>
<p class="zw-paragraph heading0"><strong>Failure Lesson:</strong> Wells Fargo's sales incentive program serves as a cautionary tale. The bank set aggressive sales goals for its employees, inadvertently fostering a confirmation bias that emphasized a high number of sales as an indicator of success. This led to employees opening unauthorized accounts for customers to meet targets, resulting in a massive scandal that severely tarnished the bank's reputation.</p>
<h2><strong>Mitigating Confirmation Bias:</strong></h2>
<ol>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Diverse Data Exploration</strong>: Encourage teams to explore a broad spectrum of data rather than focusing solely on information that confirms existing beliefs.</p>
</li>
</ol>
<ol start="2">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Cross-functional Collaboration:</strong> Involve different departments in decision-making processes to gain varied perspectives and prevent tunnel vision.</p>
</li>
</ol>
<ol start="3">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Independent Review:</strong> Establish a system of checks and balances where decisions are reviewed by impartial parties to identify potential biases.</p>
</li>
</ol>
<ol start="4">
<li>
<p class="zw-list zw-paragraph heading0"><strong>Data-Driven Insights:</strong> Leverage advanced analytics tools that objectively interpret data without being influenced by human bias.</p>
</li>
</ol>
<p>In the credit union landscape, where decisions significantly impact members' financial well-being, the influence of confirmation bias is a challenge that can't be ignored. By being aware of its existence and taking deliberate steps to counteract it, credit unions can foster an environment where data-driven insights truly lead to positive outcomes for all members.</p></div>Data Analytics in Credit Unions: Where Should They Report?https://culytics.com/blogs/bi-department-structure2023-09-11T19:27:04.000Z2023-09-11T19:27:04.000ZNaveen Jainhttps://culytics.com/members/Naveen<div><p class="zw-paragraph heading0"><a href="{{#staticFileLink}}12222130477,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}12222130477,RESIZE_710x{{/staticFileLink}}" width="710" alt="12222130477?profile=RESIZE_710x" /></a></p>
<p class="zw-paragraph heading0">In the age of information, data analytics has become an integral pillar for many businesses. This is especially true in the financial sector, where credit unions are leveraging data to make more informed decisions, personalize member experiences, and drive growth. But a question often arises: To which department should the data analytics team report?</p>
<h2><strong>Common Reporting Structures:</strong><span> </span><span> </span></h2>
<p class="zw-paragraph heading0">Across many credit unions, we see the data analytics function nestled under various departments, including:</p>
<ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>IT:</strong> Given that data handling and processing often require sophisticated technical tools, it's not uncommon for data analytics to find a home within IT.</p>
</li>
</ul>
<ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Finance:</strong> As finance departments deal heavily with numbers, forecasts, and trends, they sometimes oversee data analytics as well.</p>
</li>
</ul>
<ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Marketing:</strong> In an age where customer personalization is key, marketing teams often rely heavily on data to inform their strategies.</p>
</li>
</ul>
<ul>
<li>
<p class="zw-list zw-paragraph heading0"><strong>Operations, Strategy, Lending, etc.:</strong> These departments, while different in nature, all use data to optimize processes, strategize for the future, and make lending decisions, respectively.</p>
</li>
</ul>
<h2><strong>The Role of Executive Leadership:</strong><span> </span><span> </span></h2>
<p class="zw-paragraph heading0">While the departmental alignment might vary, one factor remains consistent: the importance of executive leadership. Any strategic initiative, including a data analytics program, requires strong backing from senior leaders. Their support is not just in terms of resources but also in fostering a data-driven culture. Often, the function to which analytics reports is influenced by which department is the most proactive and enthusiastic about leveraging data.</p>
<h2><strong>Data Analytics: A Horizontal Function </strong></h2>
<p class="zw-paragraph heading0">While it's essential to anchor the data analytics function to a particular department for clarity, it's equally crucial to understand that data analytics isn't limited to one specific domain. It's more of a horizontal function, stretching its tentacles across multiple business units and providing insights, analyses, and recommendations. So, irrespective of where it officially 'sits,' its reach should be organization-wide.</p>
<h2><strong>In Conclusion:</strong><span> </span><span> </span></h2>
<p>The ideal reporting structure for data analytics within a credit union hinges largely on the organization's unique needs, culture, and leadership dynamics. What's most crucial is ensuring that the function receives executive backing, clear direction, and the freedom to collaborate across all units. After all, in an information-driven world, data knows no boundaries.</p></div>Multi-Year Journey Through Data Transformationhttps://culytics.com/blogs/multi-year-journey-through-data-transformation2021-05-31T15:44:57.000Z2021-05-31T15:44:57.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}9012703891,RESIZE_1200x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}9012703891,RESIZE_710x{{/staticFileLink}}" width="710" alt="9012703891?profile=RESIZE_710x" /></a></p>
<p>In this competitive and data-driven world, there are more possibilities for credit unions to leverage the personal relationships with members that have helped them in achieving a high success rate over other competitors in the market. Keeping a check with the members’ data can provide CUs with the deep member understanding they need to flourish their relations.</p>
<p>Institutions that are harnessing the value of data are successful in gaining a deeper understanding of their memberships’ trends, purchasing influences, and behavior. This knowledge of data-centric operations is leading to improved financial performance, long-lasting member relationships, reduced risks, and greater member loyalty.</p>
<p>At CULytics’s Virtual 2021 summit, Amanda Paleta shared her experience at RBFCU. She has been serving RBFCU as a Metadata Analyst for 20+ years. Here are key points that are worth reading:</p>
<p><em><strong>Bring data house in-order: Realize the requirement</strong></em></p>
<p>There was heavy dependence on data for reporting. When someone asked for a specific aspect of data, it was realized that either data was not correct or it didn’t lead in the correct direction. This forced the team to go back and re-tweak the approach to get the final and correct result. As a solution, a strategy was prepared to ensure the accuracy and consistency of data, and every individual in the organization should have the same idea about the data elements. And, it was realized to put a brake on artificial intelligence and predictive analysis as it is not possible to reach the right goal if the right information is not provided. So, it is important to analyze and identify the data-related foundational requirements.</p>
<p><em><strong>Role of Analytics: To be successful</strong></em></p>
<p>It is important to check everything after the fact to understand the mistakes and areas that need to be improved. Also, try to understand how these were handled in past and what measures are required to be taken to not repeat them in the future. As a solution, it was realized that a proactive approach is required to eliminate the barriers. Other parameters were forecasting- what is coming, analyzing the market during Covid-19 and a new way of life for both an organization and members at present, after 6 months or a year. It is important to make a strategy and work on analytics to serve the best to the members in the future.</p>
<p><em><strong>Data Organization: Culture Shift and Mindset Shift</strong></em></p>
<p>It is vital to realize that data is everywhere; like in– different systems, spreadsheets, as a piece of information in people’s minds. And the requirement is to maintain this data at one central location. While organizing data, one thing that should be in mind is to not miss data at any point and do not keep any bad data. This data can be used to exponentially grow analytical and reporting power. Not just this, an organized data can help in analyzing the trends and make useful decisions.</p>
<p><em><strong>Prioritization: What needs to be Looked at First?</strong></em></p>
<p>It is not always known, what the best approach is. So, flexibility towards making amendments in actions is the key to achieve better outcomes. Working on current issues, clarify data-related issues, clean them up across the business can be helpful in building a solid foundation.</p>
<p><em><strong>Used Case Scenario: Multiple outcomes and activations</strong></em></p>
<p>The definition of one element varies from department to department. For example - Loan officers have different roles for different work scenarios. Business lending is defined differently than mortgage and consumer lending. So while doing conversations with officers, it might be difficult to figure out the right answer and to understand why analytics is not working. So, as a solution, one simple report was prepared for overall management but it was realized that there are terminologies that are unified and cleaned up in each of the lending areas and this affects the frontline staff and branches to do their work effectively.</p>
<p><em><strong>Data Governance: A part of a revolution</strong></em></p>
<p>If the steps are not taken in the right order, the right answer will not be attained. Keeping things recorded and in order is a must to move forward towards the set organizational goals. With the new regulations and compliances coming on board, data governance will be able to tell where the data came from, what it means, how it’s been used, what the intent behind it is. The regulators can see it and make a decision accordingly.</p>
<p><em><strong>Predictive Analytics: It’s important, not mandatory</strong></em></p>
<p>Every organization has different ways of working and serving its members. With predictive models, it becomes easier to understand to know the members’ needs and fulfill them. If it is not around, sometimes; members might be served with what they don’t need. So it is important, to know the members’ behavior, buying habits, etc.</p>
<p><em><strong>Communication: That can’t be ignored</strong></em></p>
<p>Communication should be at the core so that there is a successful execution of the operation. During Covid-19, many people are working from home. So, maintaining some resources, which can assist the staff in solving their doubts and queries, will be beneficial to improve the work efficiency. Techniques like Power BI can be used for this.</p>
<p>Due to covid-19, member behavior has significantly changed. They are experiencing much more digital engagement and expecting it. As a credit union, an organization has to show that it knows them better than any other company out there. So, strategies are very important in today’s competitive environment.</p>
<p>So, these were the approaches on which RBFCU worked during their <strong>data transformation journey.</strong> Strategy based approach is a must to entertain the requirements of members. For effective functionality, make flexibility, in operations, a norm. Data is important for the organization as well as member experience. Understand the organizational goals- long term and short term – and ease your data transformation journey.</p>
<p><strong><a href="https://culytics.com/articles/when-it-works-one-credit-union-s-multi-year-journey-through-data-" target="_blank">Checkout the full discussion here.</a></strong></p></div>Leveraging ACH Data to Produce Real Outcomeshttps://culytics.com/blogs/leveraging-ach-data-to-produce-real-outcomes2021-01-14T22:05:57.000Z2021-01-14T22:05:57.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8423871466,RESIZE_930x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8423871466,RESIZE_710x{{/staticFileLink}}" width="710" alt="8423871466?profile=RESIZE_710x" /></a></p>
<p>“The world is now awash in data and we can see consumers in a lot clearer ways.”- Max Levchin</p>
<p>A transformation doesn’t exist until action is taken. One thing is to have amazing data and another thing is to be able to expose that data and visualize it. Until you use it to take action and bring effective outcomes it is not a transformation; it is just a trial on data. The team at <strong><em>Sound Credit Union</em></strong> has used data, analyzed it, and made decisions based on the analysis. So, here <strong><em>Martin Walker</em></strong> is sharing the experience at the Transformation Challenge for 2020 summit organized by the CULytics. Read on to know more about the transformation journey:</p>
<p><strong>A Transformation Culture</strong></p>
<p>Transformation is all about the culture that is hard to change but not impossible. Previously, activities were performed based on credit union needs. These were uninformed and one size used to fit all. With the transformation, activities are performed based on the member needs personalized to the individual member and based on behavioral data and understanding how the members can be best served.</p>
<p><strong>Tools for Transformation</strong></p>
<p>“Without knowledge action is useless and knowledge without action is futile.” – Abu Bakr</p>
<p>Technology can be a part of the puzzle but it is a small part of the puzzle for Sound Credit Union. One of the most important parts was feedback loops and collection channels from members. Having different ways to understand, what the member is doing, what they are trying to do, and what they want to do, and having a variety of ways to get that information. The other pieces of the puzzle are <strong><em>automation and communication platforms</em></strong>. Members can’t be served individually from a marketing perspective. Serving members manually is a hard nut to crack. But using automation at this point can help in saving cost and resources. The third piece of the puzzle was data science and analytics for behavioral insights. With the help of <strong><em>information builders</em></strong>, it is easy to know the behavior of the member and made analysis accordingly.</p>
<p><strong>Wins From ACH</strong></p>
<p>According to the urban dictionary, <strong><em>ACH</em></strong> stands for <strong><em>Analytics Create Happiness</em></strong>. Here are five cases that Sound Credit Union got from ACH:</p>
<ol>
<li><strong>Govt Shutdown<br /></strong>Identified and helped specific members impacted by the 2018 government shutdown. <br /><strong>The Old Way</strong> – Previously, offer was sent to everyone, even though it is not relevant for the vast majority, or don’t send the offer at all.<br /><strong>The New Way</strong> – Now, list of members is generated with ACH payroll deposits from federal government employers. Furthermore, members are identified with a missing ACH payroll deposit from the previous pattern and reach out to impacted members and offer a 90-day, interest-free loan. <br /><br /></li>
<li><strong>CUDL Optimization<br />The Old Way</strong> – Previously, a single process was used for all transactions, regardless of use case.<br /><strong>The New Way</strong> - Upon analyzing the ACH data, it was discovered that 80% of ACH originations were for CUDL loan payments. Thus, the case-specific solutions can be deployed that quickly & significantly impact the time and labor cost involved in the activity.</li>
</ol>
<ol start="3">
<li><strong>Courtesy Pay Enhancement<br />The Old Way</strong> – Earlier, the “full” courtesy pay benefit was provided to all opted-in members, on all draft/checking shares. The result was; many members overextend themselves with courtesy pay and have difficulty in escaping the cycle.<br /><strong>The New Way</strong> – Now, courtesy pay benefits are calculated based on recent ACH deposit activity to the draft/checking share. This is helpful in observing that members-only extended a courtesy pay limit that they can likely payback based on recent activity trends. If the member has a consistent level of deposit activity then a certain level of courtesy pay benefits can be provided and if that deposit activity is impacted; the courtesy pay benefits will change in an automated manner.</li>
</ol>
<ol start="5">
<li><strong>Credit Card Offer Targeting <br />The Old Way</strong> – Before, a card offer was sent to all members who fit the qualifying credit box, typically resulting in low conversion.<br /><strong>The New Way</strong> – Now, ways can be looked to Target members with ACH payments to card companies and no sound credit card<br />The result - A 41% increase in new cards from Jan/Feb 2019 to Jan/Feb 2020</li>
<li><strong>Member View Portal<br /></strong>The data analytics team built the Member view portal. The online staff can see all kinds of information on members. One of the most important parts of the puzzle that staff can see is the ACH activity of the last 60 days. They see the recipients, etc. Also, they can see who is offered credit cards, refinance opportunities, etc.</li>
</ol>
<p><strong>Analytics Create Happiness!</strong></p>
<p>Analytics has truly helped in creating happiness for the members and the organization as it is:</p>
<ul>
<li>Proactively helping members during potentially stressful times</li>
<li>Making it easier for members to do business with us</li>
<li>Saving members’ money</li>
<li>Helping members’ end cycles of high fees</li>
<li>Providing more personalized service<br /> </li>
</ul>
<p>Analytics is useful in making the required changes in the organization, better serving the members, and creating never-ending relations. So, analyze the need of adopting analytics as it is invaluable to the process of improving and optimizing the operations.</p></div>API Lead Approachhttps://culytics.com/blogs/api-lead-approach2021-01-02T13:55:54.000Z2021-01-02T13:55:54.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8375029869,RESIZE_1200x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8375029869,RESIZE_710x{{/staticFileLink}}" width="710" alt="8375029869?profile=RESIZE_710x" /></a></p>
<p>Nowadays, the demand for data-driven business decisions is high. With more data being stored “in the cloud” by CRM systems like Salesforce, Sage CRM, etc., a solution is required to access the data in a short time and at one place. This is where APIs come into the picture. API short for Application Programming Interface allows applications to access data and co-operate with external software components. It is useful for delivering the user response to the system and the system’s response back to the user.</p>
<p>To help in understanding it better, Jordan Lehrman VP of IT at Solarity Credit Union shared his version on API Lead Approach at Transformation Challenge for 2020 summit organized by CULytics. Solarity Credit Union is a leading name operating from Washington State with 60k + members.</p>
<p>Here are key highlights from the session:</p>
<ul>
<li>Problem Definition
<p>At Solarity Credit Union, there were three big problems to deal with.</p>
<ul>
<li>Integrate Disparate Systems into a single data model – To avert confusion and for the ease of work.</li>
<li>Create a 360 Degree view of our Members – To better know the members and related things.</li>
<li>Expedite Development Time from months to days – For improving efficiency.</li>
</ul>
</li>
<li>Architecture – Old vs New
<p>Solarity Credit Union had an architecture having point to point connections. Very messy and hard to maintain and took much time to make changes. So the organization moved to 3 layer API architecture. Everything is around experience API. Salesforce Experience API is the platform that is used to provide a 360-degree view of the members. 75% of the staff spends 90% of their day in activities in which they interact with members. It is one system that the staff used to focus on what matters the most. So, this new architecture that Solarity Credit Union cooked up has about thirty systems integrated to create a dynamic API architecture. Through these different service APIs, it has become easy to go across all over different core systems and to create a single data model to get a single view of our members. And ability to take action for all these systems from one single place has been a game-changer for the organization as it allowed to focus on member interactions instead of the system that is needed to be in.</p>
</li>
<li>API based Architecture – The Benefits
<p>This API lead approach in this platform helped in saving hundreds of hours required for manual tasks, Solarity Credit Union successfully made 45 million API calls and performed 35000 transactions directly. The benefits of using API are Key Actions, Core Transactions, Contact Info Updates, Card Payments, Card Edits, ACH Setup, Loan Extensions, Pay-Off Quotes, Scheduling via O365, Email/Text/Phone, and View of info from all key systems. Moreover, APIs allow a specific user to use data more quickly, easily and efficiently when they are looking to do something specific. API is like a door or window in to the software program that empower others to interact with it without requiring the developer to share the code.</p>
</li>
</ul>
<p>So, understand the needs of your organization and make suitable decisions to improve the work experience and member engagement.</p></div>The Amazon Lending Experiencehttps://culytics.com/blogs/the-amazon-lending-experience2020-12-16T15:32:47.000Z2020-12-16T15:32:47.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8294746686,RESIZE_1200x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8294746686,RESIZE_710x{{/staticFileLink}}" width="710" alt="8294746686?profile=RESIZE_710x" /></a></p>
<p>When a company employs a data-driven approach, it means decisions are made on the basis of data analysis and interpretation. This approach enables the companies to flourish in the area of specialization and achieving the goal of better serving their customers and consumers. Technology, people, and process all work together to make a financial institution a “Data-Driven Institution”.</p>
<p>A session was conducted by CULytics in which Jaynel Christensen (VP of Lending at Commonwealth CU) shared her version of the transformation journey. The followings are the key points:</p>
<ol>
<li><strong>Every journey<strong>- small or big – starts with a challenge. For them, it was in 2017-2018, when their Unsecured Personal Loans grew 2.1% annually and Credit Card growth was .3% annually. They identified challenges with the loan application process. It was lengthy for the members and the team. </strong></strong></li>
<li>To make the transformation fruitful, Commonwealth CU started with identifying <strong>Goals and Methods</strong> to attain those goals. The aim was to find an efficient process for loan processing, making member process digitally focused and grow loans organically. But the question remains, how to find software with a quick implementation process to achieve these organizational goals and support to make changes of this level.</li>
<li>The <strong>transformation</strong> was not different and underwriting changes to be more aggressive. They modeled another CU and implemented the same software and saw success with minimal change in Delinquency and Charge Off ratios. These delinquencies were closely monitored throughout the process. Also, they aligned their internal Auto Approval model to mimic the software setup. For this, the minimum income for any offer was $ 30,000 annually. Members received multiple product offers and this created the Amazon experience; when the member needed a product; it was available for them. They had a total ratio of a maximum of 40% for any offer to be received and members with a credit score of 600 or higher received a firm offer of credit.</li>
<li>To make Data-Driven Changes, it is necessary <strong>to Review and Analyse Data Frequently</strong>. At Commonwealth CU, they did weekly decision review meetings to determine campaign criteria based on their risk cover. This helped in determining what needed to be adjusted. For example: lowering the unsecured max limit offered due to the growth during the 1st campaign. They also used Static Pool Analysis to determine the success and risk of each campaign and overall net yield.</li>
<li><strong>Challenges<strong> make the transformation journey more pleasant. For Commonwealth CU, challenges were concerned with the unsecured growth rate and future delinquency as Unsecured Portfolio balances increased over 25% from August 2018 to August 2019 and Credit Cards portfolio balances increased 12.6% over the same time period. But, the efforts gave the right result and at the end of the year, the program was maintaining delinquency and charge-off ratios within their acceptable range. </strong></strong></li>
<li>Another concern was from Examiners. How they have achieved that level of success in one year was a big question. But, they were prepared and answered well. They provided examiners a 3-inch binder full of reports on how the program was monitored throughout each campaign based on data reviewed. This binder included:
<ul>
<li>Campaign production reports</li>
<li>Deep dives into decisions</li>
<li>Static Pool Analysis for Campaign comparison</li>
<li>Net Yields by Loan Type</li>
<li>Program ROI</li>
</ul>
<p>It was impressive for examiners and they left complimenting Commonwealth for the innovative way it was servicing the members.</p>
</li>
<li>A dedicated investment of time, resources, and efforts pay off the best with great learning. At Commonwealth, they got to that the growth was based on Organizational Support of the program and Data Analytics for program refinement. With organizational buy-in and data analytics, they achieved the following results:–
<ul>
<li>Over 4,500 new loans added to the books</li>
<li>Over $40M new Original Loan Balances</li>
<li>9.6% average rate</li>
<li>Delinquency within our acceptable range</li>
<li>Charge off % well below our average charge off rate</li>
<li>Average Credit Score was 700</li>
<li>Over 7% net yield on program</li>
</ul>
</li>
<li>Besides, they learned to make changes when needed based on your data. If you have data and you don’t use it; there is no means to keep it. So, they finely tuned Total Ratio acceptance based on risk tolerance and lowered it from 40 % to 35 %.
<ul>
<li>They Limited member new loan exposure-
<ul>
<li>If a member accepts an offer in 1 campaign, they were removed from the future campaign for 6 months</li>
<li>Members can only accept 1 Secured and 1 unsecured offer per campaign</li>
</ul>
</li>
<li>They implemented a Complete Income Validation test to fine-tune appropriate minimum income-
<ul>
<li>Minimum income raised to $35,000 annually</li>
</ul>
</li>
<li>They adjusted limits to manage growth rates-
<ul>
<li>Unsecured Limits lowered by 40%; maximum limit offered $12,000</li>
</ul>
</li>
<li>They found ways to get more information on Competitor balances-
<ul>
<li>Auto Balances over $5,000 at other FI’s added</li>
<li>Credit Card Balances at other FI’s added</li>
</ul>
</li>
</ul>
</li>
</ol>
<p>Also, if the member has a mortgage, we give offers and target them for strategic loan consolidation. The journey of the Commonwealth to become a data-driven company was a long term process. The implementation of the amazon lending experience for our members through the Cunexus software was another step in the journey.</p>
<p>This was the Transformation journey of Commonwealth Credit Union. Use a visible, productive, and transformative approach to make your journey.</p>
<p>To know more about the Journey, get the <a href="https://vimeo.com/469254955" target="_blank">full video by clicking here</a></p></div>Executing Advanced Analytics Do's and Don'thttps://culytics.com/blogs/executing-advanced-analytics-do-s-and-don-t2020-12-11T14:28:49.000Z2020-12-11T14:28:49.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8274292899,original{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8274292899,RESIZE_710x{{/staticFileLink}}" width="710" alt="8274292899?profile=RESIZE_710x" /></a></p>
<p>Credit Unions are more data centric now than ever before. They are emphasizing on collecting, storing harnessing data and laying strong data-based foundations. To make decisions which are member centric, and to be able to take actions where they are needed most, Credit Unions need to be have a relevant analytics strategy and the capabilities to execute it in an efficient manner.</p>
<p>In this article we discuss the different approaches to consider and use when executing an advanced data analytics strategy and the pros and cons of these approaches. The insights confirm to the discussion of panelists, at a webinar on Executing your Advanced Analytics Strategy hosted by CULytics. The panelists included Lee Brooks, SVP, Enterprise Data Analytics, Virginia Credit Union: Ken Kondo, VP of Innovation and Software Development, Redwood Credit Union : Scott Kaylie, Data and Analytics Leader, Solarity Credit Union : Shubhankar Jain, Advisor, CULytics : and Troy Del Valle, VP, Business Intelligence, Hudson Valley Credit Union.</p>
<ol>
<li>
<p>Foundation to have before investing in Artificial Intelligence and Machine Learning.</p>
<p>Data is a tool and not the end goal. Here are some key actions you can take to build a strong foundation -</p>
<ul>
<li>Focus on four main factors-
<ol>
<li>Data warehousing</li>
<li>Business Intelligence</li>
<li>Data Science</li>
<li>Data Governance</li>
</ol>
</li>
<li>Data warehousing is very key for any predictive behavior. It helps to build a rich history which the Credit Union should build upon. Propensity Pay models, check fraud models, are some examples how advanced analytics strategy can be beneficial and drive value.</li>
<li>Bring in the right data from right resources into one data warehouse. This will help you captivate Power BI tools and take on various machine learning initiatives.</li>
<li>Evaluating use cases for more advanced analytics can provide value.</li>
<li>It is equally important to understand what particular data elements mean.</li>
<li>Engage with business functions and understand the business value.</li>
</ul>
</li>
<li>
<p>Where do use cases come from and how are they prioritized?</p>
<p>How do Credit Unions put the use cases together? Here are a few examples-</p>
<ul>
<li>Ideas from executives and reaching a consensus between the It team and C level leaders.</li>
<li>From the direction where the business is headed.</li>
<li>From servicing lines for example Human Resources Department.</li>
<li>Prescriptive and reporting requests.</li>
</ul>
<p>These use cases are prioritized based on the impact they have on business, and by being inclusive of all business functions. Analyzing the member facing impact, which gives a common language to the entire organizations.</p>
</li>
<li>
<p>How long does it take to build a data foundation? Can it be built in one day or is it a continuous effort?</p>
<p>This is an ongoing effort. With business functions, there are always going to be new data sets which means they will need to be consolidated as often as possible. One considering factor in the time it takes to build the foundation is what you are working on – for example, whether it is a machine learning model or a reporting architecture. The process can also depend on the understanding you have the data.</p>
</li>
<li>
<p>What is the technology stack you should be using? What are the pros and cons?</p>
<p>There are two ways to look at your technology stack - choose the stack which you believe that your team can manage and provide value from or provide the tools to data savvy employees. Some common tools for a technology stack include SQL, Informatica, SSRS for reporting with tableau, R as predictive modelling platform, and Tableau for self-servicing dashboards. ETL, BI tools allow you to scale. Informatica is a good tool for loading. Augmented analytics, Data IQ are other common tools. A practice of setting R file templates for machine learning, and implementing power BI for data visualizations and reporting so as to build critical thinking is a unique practice adopted by credit unions.</p>
</li>
<li>
<p>What are some challenges which have been encountered? What is the way forward?</p>
<p>Some common lessons learned by Credit Unions are as follows-</p>
<ul>
<li>Bad data will always be there. You have to come up with ways to counter that.</li>
<li>Unexpected work can derail progress towards goals.</li>
<li>Collecting the data and helping the organization develop an understanding the process can be a challenging process.</li>
<li>Not being nimble with the process is a common mistake.</li>
</ul>
</li>
</ol>
<p>Credit Unions are now moving towards using machine learning for a self-service approach, Data warehousing, and business intelligence. Building self -service dashboards, improving on data governance policies, building predictive models and implementing them effectively are other objectives.</p>
<p>CONCLUSION</p>
<p>Having business functions visualize what they can achieve with data and aligning with them can make a positive difference, and it increases transparency and data literacy. For data foundations it is important to keep evolving as you move forward. Always weigh the return when you are going after an initiative. There is always room to grow.</p></div>Lending Transformation: Old Vs Newhttps://culytics.com/blogs/lending-transformation-old-vs-new2020-12-07T15:19:03.000Z2020-12-07T15:19:03.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8258007654,RESIZE_710x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8258007654,RESIZE_710x{{/staticFileLink}}" width="626" alt="8258007654?profile=RESIZE_710x" /></a></p>
<p>COVID-19 has helped us to understand how data, analytics, and technology can simplify everything – right from placing orders online to getting door-step deliveries, having video chats with the family to running a business smoothly via digital mediums among others. Similar effects can be seen when it comes to lending in the financial industry. These transformational changes are triggered by emerging customer behavior, growing expectations, fintech disruption, newer channel proliferation, as well as the adoption of cutting-edge and innovative technologies, and rapid digitization of business and community.</p>
<p>So, for a better understanding of lending transformation; a webinar was conducted by CULytics. It is about how most progressive financial organizations are responding to challenges and making decisions to stay relevant and serve the best.</p>
<p>Jason Pieratt (Data Analysis Manager at Park Community Credit Union) hosted the sessions with the speakers - Jade Beckman (Vice President Consumer Lending at Mountain America Credit Union), Vishal Kaistha (Product Manager at First Tech Federal Credit Union), David Mitchell (Senior Manager at Portfolio Risk and Operations Alliant Credit Union), and Patrick Wiginton (Loan Origination System Officer at America’s First Federal Credit Union).</p>
<p>Followings are the highlights of the session:</p>
<ul>
<li><strong>Lending Transformation:</strong>
<p>Due to COVID, acceleration to Digital has increased as new technology and vendor partnerships increased. Let the members know what are the offers, how to access them, and accept those offers. Help them understand that things can be done digitally through mobile banking, transactions, etc. Also, get associated with the best-of-breed service providers, mobile providers, and bring in the best-in-class third-party vendors & plug in with mobile vendors to enhance the digital experience of the members. COVID-19 has helped people know digital things more and come online to perform actions that were previously completed through physical branches. This will help the management more comfortable to put things together and make plans from a strategic standpoint.</p>
<p>Some big changes like data entry can be automated and the workforce can be reserved to support and solve queries of the members and vendors to enhance the digital experience of the members. The use of mobile banking has increased due to COVID, so take things online to avoid physical interaction/ contact.</p>
</li>
<li><strong>The Lending Lifecycle </strong>
<p>The lifecycle of any element comprises of steps taken to make and maintain that particular element. For example- The loan cycle begins when a prospective borrower inquires about it, and it ends when the borrower pays off the loan. Similar is applicable for Credit Cards. Members can be incentivized to encourage the use of credit cards and make timely payments. For this, use the data and make suitable decisions. It is seen that the response rate increases when members are incentivized.</p>
</li>
<li><strong>Post-COVID Changes </strong>
<p>The COVID-19 crisis has increased the need for digital lending platforms and processes than ever before. So, to respond to the needs of the existing borrowers and future borrowers, it is required to simplify the process and offer more options to borrowers. Also, financial institutions need to take preventive measures to be in a safe place. For example- before COVID, no income verification was required or generally asked but it will be good to check in this time. Make observations based on the data. To find out the member behavior, a trial was done by observing 30, 60, and 90-day activity of members, who took the card to find out- What they are doing with the card? Results were interesting as about 80% did balance transfer that was as per expectations.</p>
<p>To bridge the gap between old ways and new ways of doing things can be considered with 2 Ps i.e. Patience and Progress. Be patient at every step of progress as transformation doesn’t happen overnight. Look at the data available and coming from the usage of products and make decisions keeping a user-friendly interface, quick initial decisioning, cloud integration, and advanced metrics at the core. Personalization can be done to meet the expectations of the members. Know the needs of the members and take measures to fulfill them to build strong and long-lasting relations. Use data, tie together data points to come to a conclusion to simplify a lot of manual work.</p>
</li>
</ul>
<p><a href="https://culytics.com/articles/webinar-lending-transformation" target="_blank">Visit Here for the Complete Webinar.</a></p></div>Data Journey: Building Strong Analytical Practiceshttps://culytics.com/blogs/data-journey-building-strong-analytical-practices2020-12-03T15:06:37.000Z2020-12-03T15:06:37.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8244843261,RESIZE_1200x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8244843261,RESIZE_710x{{/staticFileLink}}" width="710" alt="8244843261?profile=RESIZE_710x" /></a></p>
<p>Good data and data strategies allow organizations to measure and establish baselines, benchmarks, and goals to keep moving in the area of specialization. The importance of data is indispensable and to help us understand it better, CULytics organized a virtual Summit in which Ken Senus, Chief Operating Officer at St. Mary Bank explained the Data Journey.</p>
<p>Ken holds 25+ years of banking experience and oversees operational strategies to help the bank in increasing memberships and geographic presence. St. Mary’s Bank is the first Credit Union in the United States, started in 1908. It is an inviting and inclusive credit union that strives to create opportunities for individuals, businesses, and families in New England.</p>
<p>Followings are the highlights of the Data Journey of the St. Mary’s Bank to build analytical practice:</p>
<ul>
<li>The data journey started in 2014 with Data Cleaning - in preparation for the upcoming conversion effort.</li>
<li>In 2015, for Core Data Conversion- A Data Warehouse was established for reporting purposes. An interactive tool –Workbenches- is used that allows the management to look at the data they want. This eases the workload of data teams to continuously change dashboards and reports. A workbench gives a series of drop-down and filters that managers can easily look at the desired information. This way, the Loan Workbench can help in filtering the loans in a particular time period with respect to the type and other features as per the requirements. Similar is the case with the Deposit Workbench; it helps in looking at where the member is putting his/ her money.</li>
<li>In 2017, we focused on creating workbenches for teams. For this, we used interactive analytics to become more strategic. In 2014, we build a data warehouse with the help of Arkatecture. It helped in getting the desired data but a time drain of IT staff started occurring. Now, we are using Arkalytics Cloud-Based Solutions. It converted all workbenches/ dashboards/ and extracts from in-house solutions. Administration and day-to-day support activities are now done with Arkatecture. Arkalytics team is available to build custom solutions if needed. Now, our IT staff can focus on strategic jobs and rest assured for the data requirements.</li>
</ul>
<p>Also, we are using Development Sandbox. Everything that we have on the Arkalytics platform, a copy of it is available here. The plus point - we can play with things here and make changes as per the need. It helps in reducing the dependency and trying new things without modifying the actual system.</p>
<p>Now, to become more strategic, dashboards are introduced to enhance strategic member interaction. The usage of the dashboard can be understood with the followings:</p>
<ul>
<li><strong>Member Mortgage Payments:</strong> People are associated with more than one financial institution. SMB data team pulls ACH data to see if the members have mortgages outside of St. Mary’s Bank. Data information can be used for direct marketing campaigns that are very successful in gaining refinanced mortgages.</li>
<li><strong>Marketing Automation Paths from Data Warehouse: </strong>
<ul>
<li>New Member Onboarding: When a member opens a new account, he/ she receives a follow-up email informing about the account and offer them additional services at 5, 29, 59, and 89 days from the time of opening.</li>
<li>Wallet Pay:An email is triggered to any member who signs up for the wallet.</li>
<li>Maturing Auto Loans: Prior to the maturity of a member’s auto loan they receive an email inviting them to apply for the next loan and take advantage of MRC among other services.</li>
</ul>
</li>
</ul>
<p>Data warehouse not only helps business to grow but also useful to look at data & what is happening with our members. This is a great tool to look at the connected dealers, more engaged dealers, and the dealers which bring more business.</p>
<p>Also, it is useful to illustrate month-by-month activity with credit scores coming of each dealer, delinquencies, LTV, etc. Everyone has a core system that shows the 360 degree view i.e., what accounts you have, different relationships, etc.</p>
<p>VIEW THE <a href="https://vimeo.com/469483495/265f5d6a60" target="_blank">FULL SESSION HERE</a></p></div>4 Step Iterative Process : Building A Relevant Analytics Practicehttps://culytics.com/blogs/4-step-iterative-process-building-a-relevant-analytics-practice2020-11-27T14:36:06.000Z2020-11-27T14:36:06.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8220372885,RESIZE_930x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8220372885,RESIZE_710x{{/staticFileLink}}" width="710" alt="8220372885?profile=RESIZE_710x" /></a></p>
<p>What are the crucial steps towards building an analytics practice that will help your Credit Union perform in a manner which is relevant and progresses in a constantly changing business environment?</p>
<p>In this article, we throw light on how to build a relevant and successful analytics practice. The insights confirm to presentation by Michael Lindberg, Analytics and Business Consulting Services, Vice President, Wings Financial Credit Union, at the 5th Annual CULytics Virtual Summit.</p>
<p><strong>What will an analytics practice help you achieve?</strong></p>
<p>The best fit analytics practice will -</p>
<ul>
<li>Help you understand business strategy in a way that ensures clarity and focus of priorities.</li>
<li>Help you find and make data accessible in a useful manner, so that the end user is able to stay encouraged and own the data.</li>
<li>Provide you with stable and consistent definitions, so that truth is ensured within the organization.</li>
<li>Provide you with right time/ right place interaction so that membership is served in the best manner.</li>
<li>Provide you with an output that will enable your business strategy so that business ownership is driven.</li>
<li>Help you store data to provide scalability of business so that long term capabilities are ensured.</li>
</ul>
<p><strong>The 4 Step iterative Process </strong></p>
<p>The four-step iterative process to build a desired analytics practice is discussed below-</p>
<ul>
<li>Understand your value- The question that needs to be asked is, “How can you provide the best value to those you are serving?” The perspectives can be considered for business partners, organization and members. For Wings Financial Credit Union, the Wings Data Solution team defines their success as the success of business partners, and hence the solutions which are prioritized by business partners are delivered. Being aggressively collaborative, putting others first, are two practices which have proved to be very helpful.</li>
<li>Address the Pain – There are business partners and members who experience pain from time to time. The question that needs to be put forward is , “How do we relieve the pain?” For Wings Financial Credit Unions, doing report consolidations and taking frequent decisions based on the insights, working on report automations, are solutions that have helped them address the pain and provide value to members, business partners and within the organization as well.</li>
<li>Compound your Solution – Use time as an advantage. From an analytics practice perspective, it is important to engage in the same process for as long as possible so that long term benefits can be ensured. The question is, “In our organization, what solutions would set a path for success, 18-24 months from now?” In Wings Financial Credit Union, certified data models allow them to apply business definitions, financial validations, same business values etc., from which all of the data is then applied in reports, analytics, and various value providing services. This saves time and energy of having to revalidate all reports every time. If the business definition changes, it can be quickly addressed in one place. Different solutions are compounded into one or more certified models.</li>
<li>Transform your Vision- The next opportunities are not the same as the current, so as we engage with business partners, members, and our own rganization, the vision transforms with it. The analytics practice needs to move with the vision and there has to be a way to ensure consistency. For example- Wings Financial Credit Union ensures that they are delivering on a just cause. This begins with understanding and delivering member value, and then scaling the business process by using analytics to ensure that right action is taken at the right place and time.</li>
</ul>
<p>A relevant analytics practice within your organization helps you store, manage, and use updated data and take meaningful decisions which are impactful at all levels. No successful practice can be built within a day. To stay relevant, consistency is a mandate.</p></div>Building a strong Analytics Practice: Recipe for Successhttps://culytics.com/blogs/building-a-strong-analytics-practice-recipe-for-success2020-11-13T03:45:25.000Z2020-11-13T03:45:25.000ZMedhavi Singlahttps://culytics.com/members/MedhaviSingla<div><p><a href="{{#staticFileLink}}8156010672,RESIZE_1200x{{/staticFileLink}}"><img class="align-center" src="{{#staticFileLink}}8156010672,RESIZE_710x{{/staticFileLink}}" width="710" alt="8156010672?profile=RESIZE_710x" /></a></p>
<p>For all financial institutions, Data Analysis needs to be a centralized, full-time function. In the face of disruptions such as COVID-19, we should be able to expose, understand, and harness the data to enable and take action.</p>
<p>In this article, we discuss the keys to building a strong analytics Practice. The insights are taken from the presentation by Martin Walker, Vice President- Digital Experience and Innovation, Sound Credit Union, at the 5th Annual CULytics Summit.</p>
<p><strong>The Players </strong></p>
<p>To build a strong and efficient data analytics practice, the following players need to be kept in mind-</p>
<ul>
<li>The CEO – It starts at the top. If the CEO does not support the Endeavour, analytics will not be driven as a practice within the organization. Data support at the top, to use data as a factor behind decision making is crucial.</li>
<li>The Advocate – The person who is exposing the data for good. It is crucial to overcome the fear that data will impersonalize the relationship with the member. Data enables a stronger personal relationship with members.</li>
<li>The Driver- The person who decides with data, and knows how to leverage data to achieve business outcomes. Such a person is important because without action data has no impact.</li>
<li>The Interpreter – The person who speaks data and business. The interpreter helps to understand the outcome business is looking to achieve and understands how the data team can help.</li>
<li>The Champions- These are the early adopters, eager, and willing to work with data.</li>
</ul>
<p><strong>Recipe for Success </strong></p>
<ul>
<li><strong>Culture – </strong>Build a culture that supports data and enables data analysis as a practice.</li>
<li><strong>Transparency-</strong> To build a culture as described above, it is extremely crucial to be transparent in showcasing data and show how it is being used.</li>
<li><strong>Desire to be better-</strong> You should be able to feel the need to attack the disruption and improve consistently.</li>
<li><strong>Quick Wins-</strong> Irrespective of how small or big the quick wins are, it is important, for encouragement, and as a driving force, to have them under your belt.</li>
<li><strong>Find Champions- </strong>Get people on board who can adapt, learn willingly, and with excitement. They will attract more champions.</li>
<li><strong>Tell your story –</strong> You need to have someone who can tell the story and connect the dots between data and the positive outcomes.</li>
</ul>
<p>Some philosophical changes-</p>
<ul>
<li>Move from not questioning to questioning everything.</li>
<li>From trying to not fail, move towards failing fast.</li>
<li>Always believe that data enables relationships.</li>
<li>Move towards an ideology of, ‘I want everyone to know’.</li>
</ul>
<p><strong> USE CASE EXAMPLE OF SOUND CREDIT UNION </strong></p>
<p><strong>OBJECTIVE – DATA ENHANCES RELATIONSHIPS </strong></p>
<p>Sound CU had a 2-2-2 program for onboarding members. (Reaching out to members within 2 days, 2 weeks, and 2 months with particular touchpoints.) There was however difficulty in follow-through and administration.</p>
<p><strong>Solution – </strong></p>
<p>An early adopter branch-manager created a report to alert each employee to complete their next reach out.</p>
<p><strong>The Win- </strong></p>
<p>The first-year churn for new accounts decreased from 6.06% in 2018 to 4.30% in 2019, a 29% improvement.</p>
<p><strong>OBJECTIVE – DATA HELPS US BE NIMBLE, RESPONSIVE </strong></p>
<p>In late 2018, the government shutdown negatively impacted some members.</p>
<p><strong>Solution – </strong></p>
<p>The business intelligence team delivered a list of negatively impacted members to business teams to reach out with a 90-day short-term loan to help.</p>
<p><strong>The Win- </strong></p>
<p>24 loans were issued for a total of $114,200, allowing them to help members in a stressful time.</p>
<p><strong>OBJECTIVE – DATA IMPROVES EFFICIENCY </strong></p>
<p>Transaction totals that were reported by operations did not match reporting totals from other sources.</p>
<p><strong>Solution – </strong></p>
<p>The business intelligence team exposed data and worked with teams to define data and terminology, resulting in accurate reporting and common understanding.</p>
<p><strong>The Win- </strong></p>
<p>Branch staff allocation and planning were adjusted based on new reports.</p>
<p><strong>OBJECTIVE – DATA PUTS THE MEMBER FIRST</strong></p>
<p>Several business members were using courtesy pay to bridge an unpredictable gap between payables and receivables.</p>
<p><strong>Solution – </strong></p>
<p>The business intelligence team provided a list to the lending team, enabling them to extend lines of credit to members, resulting in a lower-cost, more reliable solution.</p>
<p><strong>The Win- </strong></p>
<p>For 600 business members who used Courtesy Pay, Sound CU was able to offer a solution that reduced the users’ cost by 80%.</p>
<p><strong>Learning and conclusion </strong></p>
<p>Everything we do creates data. Adoption of data analytics is not easy and many encumbrances can occur. Adoption is also critical. Ensure data quality and continue to connect data sources. It is significant to achieve some quick wins and learn from data as you go.</p></div>Three Things That Go Wrong with Data Reportshttps://culytics.com/blogs/three-things-that-go-wrong-with-data-reports2018-08-19T18:15:05.000Z2018-08-19T18:15:05.000ZRichard Joneshttps://culytics.com/members/RichardJones<div><p><span style="font-weight: 400;"><a href="https://storage.ning.com/topology/rest/1.0/file/get/72877577?profile=original" target="_blank" rel="noopener"><img class="align-full" src="https://storage.ning.com/topology/rest/1.0/file/get/72877577?profile=RESIZE_710x" width="710"/></a></span></p><p><span style="font-weight: 400;">A consistent issue most credit unions have is the lack of consistency from one report to the next. To diagnose why this inconsistency exists we should look first at these three elements of the report:</span></p><ol><li style="font-weight: 400;"><strong>Timing</strong> <span style="font-weight: 400;">- two reports run at two different end times or interval times will tell different stories. But it’s not that simple. If the report writer is aggregating reports from different sources, part of the data comes from the core and part from your loan origination system, you will find the timing between these two reports are often not in sync.</span></li><li style="font-weight: 400;"><strong>Query</strong> <span style="font-weight: 400;">- the logic used to create the reports can be different. When you are working with greater than, great than and equal to, equal to, less than and equal to, and less than, and all of the other query elements, there is a potential for error.</span></li><li style="font-weight: 400;"><strong>Definitions</strong> <span style="font-weight: 400;">- how the subject of the report is defined can change the report outcomes. Something as simple as, “What is your member number?” can be corrupted based upon the definition of a member. If one report uses a member definition of “a unique social security number above par in their membership share account” and another report defines a member as “a unique social security number” the results of the report are going to be different. This is also complicated by the reality that different software programs the credit union uses will have different names for the same elements. One software will refer to a member as a member, another will call a member an account, etc. There is a need to normalize the language differences between the variety of software providers.</span></li></ol><p><span style="font-weight: 400;">As a credit union begins the journey to utilize more data and analytics for planning an decisionmaking, it becomes necessary to conduct a series of projects designed to collect, aggregate, normalize, and make accessible.</span></p><p><strong>Collecting</strong> <span style="font-weight: 400;">data is easy and hard depending on the source. It can require special code to be written to extract the code from the software. It can require the payment of API access to the data.</span></p><p><strong>Aggregating</strong> <span style="font-weight: 400;">requires a data warehouse or a data lake to store the data. The solution a credit union chooses is not as simple as just buying a data warehouse or data lake from your core system provider or a third party, it is an intentional decision driven by the ability to the data warehouse or lake to aggregate data from most if not all of your data sources. When a $500M+ credit union has 40 or more third party data sources, this is not an easy decision and most credit unions have found the investment into a consultant that can help evaluate the options and make recommendations with a strength an weakness analysis of each of the options. There are typically several projects involved in aggregating data from these varied sources.</span></p><p><strong>Normalization</strong> <span style="font-weight: 400;">of the data needs to be part of the aggregation projects. In aggregation, the credit union is moving source data from one field with a specific name into another field in a different software. The differences in the naming protocols must be understood so we can make all of the different names for a data source consistent. If the credit union just move field A into field B, the names will not be reconciled and report inconsistencies will persist.</span></p><p><strong>Accessibility</strong> <span style="font-weight: 400;">of data does not happen automatically. To make accessibility a reality, the credit union must have a tool or a domain that allows data to be accessed across the organization. But accessibility also requires that the data, queries, and definitions be governed. A credit union has checks and balances on their deposits and cash flow, they also need checks and balances on how data is being extracted and reported upon. If we just give “unfettered” access to data without these checks and balances, there is a significant risk in reports being generated with the sole purpose of supporting a predetermined outcome, the trustworthiness of data is lost.</span></p><p><span style="font-weight: 400;">Data’s primary roles in a credit union are to:</span></p><ol><li style="font-weight: 400;"><span style="font-weight: 400;">Understand what has happened</span></li><li style="font-weight: 400;"><span style="font-weight: 400;">Understand why it happened</span></li><li style="font-weight: 400;"><span style="font-weight: 400;">Understand what can be done to change the future</span></li><li style="font-weight: 400;"><span style="font-weight: 400;">Understand how data can help to predict the future (artificial intelligence - AI)</span></li><li style="font-weight: 400;"><span style="font-weight: 400;">Understand how data can help design the future (machine learning)</span></li></ol><p><span style="font-weight: 400;">For credit unions to be successful in the future, they need to see data as an organizational asset that needs to be collected, normalized, analyzed, and accessible. This aspirational goal takes time, focus, discipline, and commitment.</span></p><p> </p></div>How much is too much personalization?https://culytics.com/blogs/how-much-is-too-much-personalization2017-10-19T13:58:28.000Z2017-10-19T13:58:28.000ZJennifer McGinnhttps://culytics.com/members/JenniferMcGinn<div><p><b><a href="http://storage.ning.com/topology/rest/1.0/file/get/2161365?profile=original" target="_self"><img src="http://storage.ning.com/topology/rest/1.0/file/get/2161365?profile=RESIZE_1024x1024" class="align-full" width="750" height="144"></a></b></p>
<p>With the vast amount of banking options available today, it is no surprise that consumers want everything in real-time, personalized to their every demand and if they don’t get it, they simply move on to the next institution that can.</p>
<p>But what price are people willing to pay to get this personalized service and instant satisfaction from their financial institutions? How much of their personal information are they willing to share?</p>
<p>The only way financial institutions can respond to this demand and personalize the customer journey is to know as much as they can about you: your habits, preferences, relationships. They even try to anticipate wants and needs – maybe suggest a retirement savings for anyone turning 18 or recommend a 30-year mortgage every time someone posts a picture of an engagement on social media.</p>
<p>So, are consumers really willing to give up all that data just to have a better experience? It seems they are. They freely post pictures, status updates and opinions on social media and websites, and they openly share demographic data leaving behind a digital footprint heard round the world.</p>
<p>This comes down to consumers wanting to be treated like individuals. And wanting to have unique experiences. </p>
<p> </p>
<p><b>The savvy consumer</b></p>
<p>Today’s consumers are savvy about their data. They know when they share direct or via social networks, that data will be used – both targeted at them individually and at an aggregate level for segmentation. They need to know they are in some level of control and that the data will be used for good not evil. And for the most part, they trust that it is secure.</p>
<p>Now the tricky part: how can financial institutions take advantage of this free sharing of data. How can they sift through the vast amounts of data to find trends, connections, and indicators to provide the right level of personalization? As more unstructured data from social media, video, images, webchats, geospatial data and others gets dumped into a data lake such as Hadoop, the ability to gain insight slips away. What once was a problem of not having enough data to identify trends or properly do segmentation has turned into a situation of having too much information to make any sense of it.</p>
<p> </p>
<p><b>Customer 360</b></p>
<p>The answer lies in the ability to synthesize all data together into an <a rel="nofollow" href="http://allsight.com" target="_blank">intelligent customer 360</a>. This is different than the traditional master data management profiles generated over the past 10+ years which were limited to structured data and exact matches. Now the true hidden nuggets of insight lie in fragments of data that can be stitched together.</p>
<p>If marketers, customer service reps and personal bankers have a better understanding of customers – identify high net worth individuals, recognize risk of customer churn, provide targeted offers – then personalization will be done right. And banks will grow and retain the right customers.</p>
<p></p>
<p> </p>
<p> </p></div>