Overview
Retaining existing customers is crucial for a business’s revenue growth, alongside acquiring new ones. Our client, a market leader, was experiencing a decline in its market share as it lost large portfolio customers to competitors. The leadership team recognized the need to identify the underlying cause and adopt a more customer-centric approach to reverse these trends and sustain its competitiveness.
The Challenge
While making corporate growth, its internal processes, procedures, communication and collaboration in between departments became more corporate focused rather customer centric which often led to conflicting priorities and a disjointed approach to customer services and experience.
To address these challenges, the company already had embarked on a comprehensive customer-centricity program where it was collecting Survey Scores, Variety of Client Support Requests such as Inventory Copy, Contract Copy, Pricing and complaints etc. It becomes more challenging along with variety of datasets along with operational datasets such as Invoices, Work Orders, Inventory etc to corelate and identify factors behind loosing customers.
Our Approach
To initiate the program, the first step involved conducting a thorough evaluation of the entire customer journey, starting with the initial contact and extending to post-sale support. Simultaneously, the team of subject-matter experts examined the available data assets in order to identify the relevant datasets for this exploratory study. For the study, the following datasets were deemed crucial by the expert panel.
- Survey Scores
- Client Inventory Request
- Client Contract Request
- Invoice Credits History
- Recurring Revenue Trend
- Work Orders Delivery
- Price Renewals and Rollbacks
- Industry Prices
The Data engineers played a crucial role in extracting, transforming, and normalizing datasets for machine learning purposes. Simultaneously, the data analysts focused on visualizing and evaluating model results and forecasts in collaboration with the business team. Through their efforts, the data analysts were able to identify customers at risk (categorized as Low, Medium, or High) based on their recurring and non-recurring revenue portfolios. This information allowed for prioritized customer retention strategies.
As a result, the leadership team gained a heightened awareness and the ability to isolate customers based on the size of their business portfolio and their retention risk (High, Medium, or Low) at an early stage. This empowered the leadership team to mobilize operations, sales, and customer support VPs, fostering collaboration to promptly address customer concerns and mitigate risks.
Leadership team became more aware and empowered to isolate customers by size of business portfolio and retention risk by High, Medium and Low at quite early stage which empowered leadership to mobilize operations, sales and customer support VPs to collaborate and address customer concerns at early stage.
Outcomes
Thanks to the successful implementation of these measures, our client achieved an impressive customer retention rate of 70%, with a particular focus on retaining customers within the large and medium business portfolios. The machine learning models played a crucial role in identifying these customers as being at risk, enabling targeted efforts to mitigate the potential loss and ultimately contribute to the high retention rate.
From the outset, the business function heads (SVPs, VPs) became more proactive in identifying potential dissatisfied customers and the reasons behind dissatisfaction, the team came to know they were falling behind in customer expectations, including slow work orders delivery, poor aggressive price revisions resulting in credits, and a lack of empathy and understanding.
Next Steps
The leadership team made a strategic decision to integrate various business applications with Machine Learning APIs as end-points in order to assess customer health efficiently and respond promptly. They achieved this by harnessing the power of machine learning and data science-based APIs.