Using the Observe.AI Conversation Intelligence Platform and its QA automation features, ServiceFirst achieves a 77% ROI and finds over $10,000 in annual savings.
ServiceFirst is a world-class contact center solutions provider. But with 2,000 agents spread across three countries, it struggled with its manual QA process. Customer calls averaged 15 to 20 minutes while their manual quality assurance (QA) process averaged 30 to 45 minutes, which meant there was only enough time to conduct two evaluations of each agent’s performance per month. This made it difficult for ServiceFirst to gain insights into customer interactions to improve customer experience and sales metrics, and help achieve revenue targets.
Compliance monitoring is also a crucial component of ServiceFirst’s QA efforts, because many of its lead generation and customer support clients are lending companies based in the United States (U.S.). As its Global Head of Speech Analytics, Jon Reynel Lumongsod notes, U.S. financial institutions must adhere to various regulatory requirements.
“Compliance failures in contact centers are common and often overlooked,” he says. “So, one of our top priorities is call monitoring to detect and address compliance errors.”
Faced with QA limitations, ServiceFirst’s speech analytics team turned to Observe.AI for a solution.
Using Observe.AI Auto QA, the speech analytics team accelerated its QA process and scaled its operations. It fully automated agent evaluations through Moments, which are key instances in customer interactions that reveal insights and trends. Such instances include determining whether an agent established a right party contact (RPC).
“We wanted to increase our percentage of RPC, which means successfully reaching the intended recipient in a call, versus our total calls, because that’s where we get conversions,” says Lumongsod.
By integrating its QA form into the Observe.AI platform, ServiceFirst leverages Moments to see correlations between specific call quality criteria and the company’s key performance indicators. Lumongsod explains, “Each of these Moments is tied to specific items on the QA form, so we can see if our agents have covered each required item during the call.”
Supervisors now assess nearly all customer interactions, thanks to automation. The first phase of this effort resulted in a 100% gain in evaluation time efficiency and an 1,800% increase in assessments since implementation.
By leveraging its new QA capabilities, the team achieved a 44% improvement in internal and external compliance, including regulatory and industry standards.
Supervisors previously evaluated customer calls by selecting them from a random pool per agent. Now, they can specifically capture compliance-related interactions, such as calls involving compliance spiels, verification, and recorded disclosures.
“By using Moments, we can identify and validate these calls and have a more targeted pool for evaluation,” says Lumongsod.
When alerted to compliance issues, supervisors can provide coaching focused on mitigating specific risks in agent performance. This proactive approach helps agents handle scenarios more effectively while reducing compliance risks.
Having full visibility into customer interactions enabled ServiceFirst’s speech analytics team to uncover behavioral issues among agents, including voicemail camping. This refers to agents remaining on an idle line or calling in to a voicemail without speaking for extended periods, leading to inefficiencies, poor customer service, and missed revenue opportunities.
“If an agent can’t reach a customer, they’re trained on proper voicemail handling, including what to say, and they end the call promptly to be ready for the next one,” explains Lumongsod. “But some agents misuse this by staying on the line unnecessarily.”
QA managers were able to identify which agents were engaging in voicemail camping, and how frequently, by scaling the interactions they captured and evaluated. This enabled them to quantify lost hours into dollar figures and address this behavior through targeted coaching. Over three months, voicemail camping instances declined by 43%.
The team also reduced call hang-ups by analyzing Moments around common drop-off points and concerns among loan customers.
Previously, ServiceFirst’s contact centers had high percentages of hang-up calls. By focusing on reducing controllable opportunities—instances where an agent can influence the outcome of a call but fails to do so—from callbacks, the hang-up rate fell by 39% and 20% at two contact centers over four months. The RPC conversion rate at these centers—the percentage of successful loan conversions after contacting the correct person—increased in tandem by 44% and 36%, respectively.
These improvements are flowing through to higher loan volumes generated for clients. “If these continue throughout 2024, we expect to generate an additional annual loan amount increase of $1,726,500 for a headcount of 15. That’s $115,100 in extra revenue per agent,” says Lumongsod.
Similarly, improvements in agents offering callbacks and adhering to the callback policy at the Alabang, Philippines site led to a 6% increase in RPC conversion rates. With 121 agents, this translates to a potential additional loan volume exceeding $3 million from May to December 2024.
Overall, the team’s automation efforts are set to deliver an additional $3.7 million in loan volume by the end of 2024 through reduced hang-up rates,improved callbacks, and RPC conversions. These efforts are also expected to result in more than $10,000 in annual savings.
Looking ahead, ServiceFirst’s speech analytics team will focus on refining rebuttal scripts and statements to help agents turn calls into loan conversions. With full visibility of interactions, the team can influence behavioral changes by training agents to effectively recognize and capitalize on opportunities to respond to customer objections.
Agents previously often failed to provide effective rebuttals to objections, resulting in missed sales opportunities. “By focusing on Moments that track agent rebuttals, we aim to improve how agents handle objections and increase conversion rates,” says Lumongsod.
Book a demo today or explore the new features of Observe.AI to enhance your operations.
.jpg)
.png)
.png)
.png)
