Using Actionable Analytics to Improve Customer Service while Reducing Costs

A short while ago, we were engaged by a large telecommunications client to address a costly claims management problem. The company receives several million billing disputes a year from its commercial customers through a variety of channels including email, web self service, customer contact centers, and batch files. Each of these disputes results in a claim that needs to be investigated and resolved. The resolution can be either in favor of the customer by issuing a corresponding billing credit or in favor of the company by issuing a rejection with an explanation. It was costing the company $70 to process each claim and, more importantly, the long time it took to resolve each claim was negatively affecting customer satisfaction.

To determine what was causing the claim resolution to be so costly and time consuming, we conducted a series of discovery sessions with the customer service representatives to observe their work, analyze the claim investigation process, and identify root causes. What made the claim management process complex was the fact that investigating a claim meant pulling large data sets from 22 different systems for billing, order management, and service assurance and correlating it to determine the validity of the claim. The company’s previous attempt at addressing this problem by providing a single portal across all 22 systems yielded no improvement whatsoever in the claim processing time or cost.

Armed with this information, we took a different approach to the problem: rather than focusing on the symptoms and trying to make small incremental improvements in Customer Service Representative productivity, we explored the possibility of automating the claim resolution process leveraging Actionable Analytics.

In order to achieve this goal, we leveraged the 5 Whys approach to root cause analysis which resulted in identifying a series of algorithms that were used to determine the validity of a claim.

The solution involved several stages:

1. Ingest the claims from the variety of channels in a highly scalable manner
2. Retrieve pertinent data sets from the appropriate systems depending on the nature of the claim. This required complex integration with the 22 systems involved combined with rules based orchestration of data retrieval based on the nature of the claim.
3. Run the Analytics Algorithms against the large data sets to determine if the claim can be resolved automatically. The algorithms combined hard rules about claim resolution with soft rules about the nature of the relationship with the customer.
4. For any claims that could not be resolved by the Analytics Algorithms, proactively create a case with all the relevant information and assign it to the appropriate Customer Service Representative queue for further investigation.

By leveraging this approach, we were able to resolve 35% of incoming claims automatically resulting in a significant reduction in claim resolution time from 30 days to a few hours. In addition, the proactive correlation of pertinent claim information before assigning a claim to a representative resulted in a reduction of claim investigation time by about 80%, increased the accuracy of the claim resolutions, and reduced the percentage of new complaints related to the same issues significantly. The solution further involved proactive management of case load to ensure proper prioritization, assignment and escalation so that claims can be resolved as fast as possible. The combination of these capabilities resulted in a reduction of claim processing cost from $70 to under $10 while achieving a commensurate improvement in customer satisfaction.

Engage the Dante team to help you harness the power of Actionable Analytics.

---