Case Study: Helpline

01 Challenge

For credit and debit card issuers, one of the most important services is fraud prevention. A group of Helpline agents are assigned specifically for this task, which demands careful analysis and usually involves direct contact with the cardholder.

When the case management system issues a suspicious transaction alert, the fraud mitigation process involves analysis of the transaction, customer contact and care, wrap-up and documentation. In most cases, as part of the investigation, a Helpline agent contacts the cardholder for clarification of the transaction. If the customer does not answer after three tries, then the agent temporarily blocks the card. An email then has to be sent to the customer about the suspicious transaction and asking them to contact the relevant Helpline representative.

At this stage, either the agent has spoken with the cardholder to clarify the potentially fraudulent transaction or an email has been issued requesting an immediate response from the customer. The agent must then undertake a series of data entry actions as part of the case wrap-up phase, including logging a ticket with the case details, recording the outcome of the analysis, filing a claim request when necessary, updating the database, and noting other information for case management records.

According to Helpline service level agreements, investigation of any suspect transaction must be initiated within five minutes of the alert. In practice, however, agents were under too much pressure handling the data entry for their individual cases to meet the demands of the SLAs. Average handle times were relatively high and agents felt their time was being used ineffectively.

In order to ensure transaction security for their clients, Helpline needed to reduce repetitive, albeit necessary, manual tasks interfering with the core activity of fraud prevention agents.

“This solution gives the agents a feeling that their time is valued more.”

Francesco Andreani

Contact Center Manager
Helpline

02 Solution

In addressing the challenge of getting the most out of the time spent by its agents in the security management line of business, Helpline adopted NiCE Robotic Automation and NEVA Assist (Attended Automation).

With NEVA Attended Automation in place, the agent has a real-time assistant providing key information and making sure all bases are covered during the case analysis stage. When an initial suspicious transaction alert is triggered, it is accompanied by an NEVA Assist pop-up window indicating to the agent the reason behind the alert (for example, an unexpected transaction in a remote country, a withdrawal over a certain amount of money, etc.).

As noted, the agent analyzes the alert and, if necessary, attempts to contact the cardholder for clarification. Another NEVA Assist window then asks the agent to confirm if the customer has made a reimbursement claim. If the customer cannot be reached, then the card has to be blocked and the customer emailed.

This is where NEVA Unattended (Unattended Automation) takes over, handling all of the mundane and repetitive elements of the wrap-up phase described above, where human intervention is not necessary (i.e., the “red tape” of case documentation). Depending on the feedback from the customer, or lack thereof, the robot will block the card, send an email, or create a claims letter.

The solution automatically updates a temporarily blocked credit card database, checking the real-time status against customer replies (i.e., “okay to process payment” or “this is not my transaction”, or no reply yet). A robot also checks the amount of time left until a blocked card is to be reactivated. The list of cards whose suspension is about to expire is then dispatched to a supervisor.

NEVA Unattended completes the entire wrap-up process, documenting the case and its outcome. The system then sends an email to the relevant agent indicating if the wrap-up was successful or not.

More time for customer care

With the implementation of NEVA Assist during the transaction analysis phase and Robotic Automation of the wrap-up, Helpline eased the pressure on agents and improved their performance.

Wrap-up phase handling times were dramatically reduced, going from 92 to 39 seconds on average for the simplest cases. More complex cases could be wrapped up in under five minutes, as opposed to eight before Robotic Automation.

This has given agents more time for customer care activity, such as cross-checking the case within the fraud prevention team and the like. The result has been increased accuracy in the analysis and more satisfied customers.

Today—and tomorrow

Today, with eight active virtual robots, Helpline handles 8,000 alerts per month with over 99% accuracy in preventing fraudulent activity. And the company now meets its process SLAs fully 100% of the time.

Looking ahead, NiCE Robotic Automation’s virtual infrastructure ensures maximum flexibility and high scalability. As Edoardo Fabris, IT Manager for Helpline, put it: “Adding a robot is far simpler than training [employees] how to perform a process.”

Helpline intends to expand the use of NiCE Robotic Automation to reduce overall average handle time for all relevant customer interactions. The company has plans to implement the solution in other areas of business activity, as well, including all its back office and offline activities.

“We envision a future in which our agents only take care of [value-generating] activities,” said Francesco Andreani, Contact Center Manager for Helpline. “We regard NiCE Robotic Automation as a powerful tool in order to fulfill this vision.”

“[Agents] can better focus on the customer, instead of focusing on manual tasks.”

Sonia Ottogalli Zanin

Team Leader
Helpline