Client Health Score Dashboard: Predictive Attrition Reduction (8%)

Client Health Score Dashboard: Predictive Attrition Reduction (8%)

Executive Summary

Reeves Institutional Advisors (RIA), faced the challenge of reactive client attrition management, leading to an average loss of $5 million in Assets Under Management (AUM) per quarter. Golden Door Asset partnered with Reeves to develop a proprietary client health score dashboard that aggregated data from multiple sources, enabling proactive identification of at-risk clients. This intervention resulted in an 8% reduction in client attrition over one year, directly preserving approximately $16 million in AUM.

The Challenge

Reeves Institutional Advisors, a boutique RIA managing over $5 billion in AUM, struggled with client attrition despite a strong track record of investment performance. Historically, identifying clients at risk of leaving was a reactive process, relying heavily on lagging indicators such as withdrawn assets or negative feedback received during annual reviews. By the time these signals surfaced, it was often too late to effectively intervene and prevent the client from moving their assets elsewhere.

The primary challenges Reeves faced included:

• Lack of Proactive Insights: The firm lacked a centralized system for monitoring client engagement and identifying potential dissatisfaction early on. Client interactions were tracked inconsistently across different advisors, making it difficult to spot emerging trends.
• Data Silos: Client data was scattered across multiple platforms, including Salesforce (CRM), their portfolio management software (Advent Axys), and survey results collected through SurveyMonkey. This made it time-consuming and cumbersome to gain a holistic view of each client's relationship with the firm.
• Reliance on Anecdotal Evidence: Advisors often relied on their gut feelings or anecdotal evidence to assess client satisfaction. This subjective approach was prone to biases and inconsistencies, leading to missed opportunities for proactive engagement.
• Inability to Predict Attrition: The firm struggled to accurately predict which clients were most likely to leave. This made it difficult to allocate resources effectively and prioritize outreach efforts to at-risk clients. As a result, Reeves was experiencing an average attrition rate of 5% annually, translating to approximately $250 million in AUM at risk each year. This attrition resulted in a loss of approximately $250,000 in annual revenue, based on their average advisory fee of 1%.
• Inefficient Resource Allocation: The lack of a clear understanding of client health led to inefficient allocation of advisor time and resources. Advisors spent too much time servicing clients who were already highly engaged, while neglecting those who were silently disengaging.

The cost of inaction was significant, not only in terms of lost revenue but also in terms of reputational damage and the opportunity cost of acquiring new clients. Reeves needed a more proactive and data-driven approach to client retention.

The Approach

Golden Door Asset partnered with Reeves Institutional Advisors to develop a client health score dashboard designed to address these challenges. The approach was multifaceted, encompassing data integration, predictive modeling, and actionable insights.

1. Data Integration and Centralization: The first step was to integrate data from Reeves' disparate systems into a centralized data warehouse. This involved establishing secure connections to Salesforce, Advent Axys, and SurveyMonkey, and developing automated data pipelines to extract, transform, and load (ETL) data on a daily basis.
2. Definition of Client Health Indicators: Working closely with Reeves' management team, Golden Door Asset identified key indicators of client health. These indicators were categorized into three main areas:
   • Engagement: Measured by factors such as the frequency of client meetings, participation in webinars and events, website logins, and interaction with online content.
   • Portfolio Performance: Assessed based on factors such as portfolio returns relative to benchmarks, risk-adjusted returns (Sharpe Ratio), and the consistency of investment performance.
   • Communication: Tracked by the frequency of phone calls, emails, and personalized communications between advisors and clients. Sentiment analysis was also conducted on client emails and survey responses to identify potential concerns.
3. Development of the Client Health Score Algorithm: A proprietary algorithm was developed to calculate a client health score based on the weighted average of the identified indicators. Each indicator was assigned a weight based on its perceived importance and predictive power. The algorithm also incorporated decay functions to give more weight to recent activity and de-emphasize older data. For example, a recent negative survey response would have a greater impact on the score than a similar response from several months ago.
4. Dashboard Design and Visualization: Using Tableau, a visually appealing and intuitive dashboard was created to display the client health scores. The dashboard provided advisors with a clear and concise overview of the health of their client relationships, allowing them to quickly identify at-risk clients and prioritize their outreach efforts. The dashboard also included drill-down capabilities, allowing advisors to explore the underlying data and understand the specific factors contributing to each client's score.
5. Proactive Intervention Strategies: Golden Door Asset worked with Reeves to develop a set of proactive intervention strategies tailored to address the specific concerns of at-risk clients. These strategies included:
   • Personalized Outreach: Advisors were encouraged to reach out to at-risk clients with personalized communications, addressing their specific concerns and offering solutions.
   • Proactive Portfolio Reviews: For clients with concerns about portfolio performance, advisors conducted proactive portfolio reviews to discuss investment strategies and adjust asset allocations as needed.
   • Enhanced Communication: Advisors increased the frequency of communication with at-risk clients, providing regular updates on market conditions and portfolio performance.
   • Exclusive Events and Webinars: Clients were invited to exclusive events and webinars designed to enhance their understanding of financial planning and investment management.
6. Continuous Monitoring and Improvement: The client health score dashboard was continuously monitored and refined based on feedback from advisors and ongoing analysis of client attrition patterns. The algorithm was adjusted as needed to improve its predictive accuracy and ensure that it continued to provide actionable insights.

Technical Implementation

The technical implementation of the client health score dashboard involved a combination of data integration, data processing, and data visualization technologies.

• Data Integration: Data was extracted from Salesforce using the Salesforce API, from Advent Axys using SQL queries, and from SurveyMonkey using their API. Secure connections were established to each data source, and automated data pipelines were developed using Python scripts and the Apache Airflow orchestration platform.

• Data Processing: The extracted data was transformed and cleaned using Python libraries such as Pandas and NumPy. Data cleaning steps included removing duplicates, handling missing values, and standardizing data formats. Sentiment analysis was performed on client emails and survey responses using natural language processing (NLP) techniques. Specifically, the VADER (Valence Aware Dictionary and sEntiment Reasoner) lexicon and rule-based sentiment analysis tool was implemented.

• Client Health Score Calculation: The client health score was calculated using a weighted average formula. Each indicator was assigned a weight based on its perceived importance and predictive power. The weights were determined through a combination of expert judgment and statistical analysis. The formula for calculating the client health score was:

  Client Health Score = (Engagement Score * Weight_Engagement) + (Portfolio Performance Score * Weight_Portfolio) + (Communication Score * Weight_Communication)

  Each of the individual scores (Engagement, Portfolio Performance, Communication) was calculated based on a series of sub-indicators. For example, the Engagement Score was calculated as:

  Engagement Score = (Meeting Frequency Score * Weight_Meeting) + (Webinar Attendance Score * Weight_Webinar) + (Website Login Score * Weight_Website)

  The weights were initially set based on expert judgment and then iteratively refined based on the model's performance. For example, if website logins proved to be a strong predictor of retention, its weight would be increased.

  A decay function was applied to each indicator to give more weight to recent activity. The decay function was exponential, with a half-life of 90 days. This means that the impact of an indicator would be reduced by 50% after 90 days.

• Data Visualization: The client health score dashboard was developed using Tableau. The dashboard included a variety of visualizations, such as bar charts, line graphs, and heatmaps, to display the client health scores and the underlying data. Interactive filters allowed advisors to drill down into the data and explore the specific factors contributing to each client's score. The dashboard was designed to be user-friendly and intuitive, with clear and concise labels and visualizations.

• Database: All data was stored in a PostgreSQL database, chosen for its reliability, scalability, and support for advanced data analytics. The database schema was designed to efficiently store and query the large volumes of client data.

• Version Control and Deployment: The code for the data pipelines and the client health score algorithm was managed using Git version control. The dashboard was deployed on a secure Tableau Server, with access restricted to authorized users.

Results & ROI

The implementation of the client health score dashboard had a significant positive impact on Reeves Institutional Advisors' client retention rates and overall financial performance.

• Reduced Client Attrition: The client health score dashboard enabled Reeves to proactively identify and engage with at-risk clients, resulting in an 8% reduction in client attrition over one year. This translated to a decrease in the annual attrition rate from 5% to 4.6%.
• Increased AUM Retention: The reduction in client attrition resulted in the preservation of approximately $16 million in AUM. This was calculated based on the initial $250 million at risk (5% of $5 Billion AUM) multiplied by the 8% reduction in attrition ($250,000,000 * 0.08 = $20,000,000 at risk of leaving), and a conservative estimate that 80% would have left, or $16,000,000.
• Increased Revenue: The increase in AUM retention directly translated to an increase in revenue. Based on Reeves' average advisory fee of 1%, the $16 million in AUM retained generated an additional $160,000 in annual revenue.
• Improved Client Satisfaction: Proactive engagement with at-risk clients led to improved client satisfaction. Client satisfaction scores, as measured by annual surveys, increased by 15% among clients who were identified as at-risk and received personalized outreach.
• Enhanced Advisor Productivity: The client health score dashboard enabled advisors to prioritize their outreach efforts and focus on clients who were most at risk of leaving. This resulted in a 20% increase in advisor productivity, as measured by the number of client interactions per week.
• Quantifiable Return on Investment (ROI): The total cost of developing and implementing the client health score dashboard was approximately $50,000. The increase in revenue of $160,000 generated in the first year resulted in a quantifiable ROI of 220%.

Key Takeaways

1. Proactive client retention is crucial for sustainable growth. Don't wait for clients to express dissatisfaction. Implement systems to identify at-risk clients before they churn.
2. Data integration is essential for a holistic view of client health. Break down data silos and consolidate information from multiple sources into a single, unified dashboard.
3. A well-defined client health score can be a powerful tool for predicting attrition. Develop a scoring system that incorporates key indicators of engagement, portfolio performance, and communication frequency.
4. Personalized outreach and proactive intervention can significantly improve client satisfaction and retention. Tailor your engagement strategies to address the specific concerns of at-risk clients.
5. Continuously monitor and refine your client retention strategies. Track the effectiveness of your interventions and adjust your approach as needed to optimize results.

About Golden Door Asset

Golden Door Asset builds AI-powered intelligence tools for RIAs. Our platform helps advisors identify risks, predict outcomes, and deliver personalized service at scale. Visit our tools to see how we can help your practice.