Concentrix Corporation, Tempe, AZ USA: Concentrix Saves Client Millions in Uncollected Premiums


Gold Stevie Award Winner 2017, Click to Enter The 2018 Asia Pacific Stevie Awards

Name of Organization / Company: Concentrix Corporation, Tempe, AZ USA:
Category: Award for Innovative Management in Financial Industries - More than 100 Employees

Entry Title: Concentrix Saves Client Millions in Uncollected Premiums


Concentrix Saves Client Millions in Uncollected Premiums

Our client is one of India’s leading insurance services institutes. We handle back end processes including claims renewals.


Customers opt for monthly, quarterly, half-yearly or yearly payment options on premiums. Every year about 2.5 million insurance policy renewals fall due. Before we implemented our tool, our client was losing a lot of these clients due to lack of persistence. Because of the volume, we estimated an improvement of 1% in this area could translate to an extra premium of nearly $3.5 million in a financial year.
Several factors fueled the lack of renewals. The system relied heavily on manual reminders to encourage customers to renew their policies. Too much of the burden of remembering and renewing landed on the customers, which resulted in many customers surrendering their policies. In addition, the client needed more individualized information about specific policyholders and the factors that led to them lapsing their policies.

We knew that a strong solution would require excellent analytical modeling to recover the lost revenue and improve the business strategy. We also knew with good execution of daily strategies and organizational management that played off our strong teamwork and Omni-channel engagement.


Organization Management

For the first part or our solution we separated customers into buckets based on their risk of defaulting. Some of the buckets contained customers who needed an action from our client in order to renew. Some required actions from our staff. And others were customers who needed to take action on their own in order to retain their policies with the client.

We developed a propensity to pay model for every different bucket, based on which customer type, where the customers were within their grace period, past behavior and which type of policy the customer had. Once we had this information, we developed strategies that included intervening with customers on multiple channels (voice, text or email) and following up several times based on risk of default. The model automated much of what had been a manual process and increased efficiencies.

Employee Development

We also realized that staff development was a crucial part of the program. Team members who did well in different segments were groomed to take on new responsibilities such as subject matter experts, call quality monitoring and work flow management. This built a stronger foundation for our team as process knowledge expertise was shared between areas. It also improves morale, which aided retention, which further aided our forward momentum in process improvements.


Our propensity to pay model increased customer policy retention by 1.07%. This resulted close to our estimate with more than $3 million savings. We reduced the lapse rate by 8% and reduced the policy surrender rate by more than 40%. We also managed to proactively retain customers from less than 2% to more than 6%.

Business Recovery

Our baseline for collections on policies which were coming due was 71.52%. We increased this to 80.99% for fiscal year 2015-2016. Diligence in following up on missing payments improved our client’s bottom line. In the first year (FY2014-2015) we found $23.13 million in missing unpaid premiums and an additional $2.24 million in the following year (FY2015-2016).
Our surrender rate dropped from 8.74% in fiscal year 2015-2016, saving an estimated $50.30 million.

We have been actively encouraging customers to use electronic funds transfers for electronic clearing services (ECS). Our campaign led to 54% ECS cases in the fiscal year 2015-2016, up from 50% in the previous year.

Supporting Links: