Misalignment between Dialpad’s support KPIs and its support value of “doing the right thing”
Agents felt too much pressure to focus on ticket quantity—instead of quality—which lessened their ability to empathize with customers
Reimagining agent performance metrics to encourage quality over quantity, while balancing the need for operational efficiency
5% increase in CSAT and a 6% increase in agent performance
Dialpad is a cloud-based communications platform that’s powered by artificial intelligence. More than 70,000 companies, including HubSpot, Uber, and Stripe, rely on Dialpad’s products to elevate web conferencing, sales calling, and call center management.
One of Dialpad’s values is doing the right thing for customers, which is why the company offers 24-7 live support across a variety of channels. Dialpad’s QA team exists to help agents live up to this ideal while maintaining their operational goals.
Analyzing voice intelligence data from its own call center product led Dialpad’s QA team to notice gaps in the customer experience. Although CSAT was relatively strong—hovering around 85%—agent empathy represented an opportunity for improvement.
“Our research found that we were facing challenges with agents not always practicing the right level of empathy,” said Brynn Gaeta, Team Lead, Support QA and Development at Dialpad. “We also found that agents didn’t feel supported to ask questions when they didn’t know the right answer.”
After further investigation, CX leadership recognized the underlying problem: misalignment between Dialpad’s support metrics and its company values.
Historically, agent performance was based on an even weighting of several operational (# of live touches, First Call Resolution, and # of solved tickets) and call center quality metrics (QA score, responsiveness, and CSAT). At face value, this seemed like a logical (and fair!) way to measure agent performance. In reality, agents felt too much pressure to deliver on ticket volume at the expense of soft skills, such as empathy. 😩
“While operational efficiency is very important, there needs to be a balance between operations and quality,” Gaeta said. “We were measuring agents more on the amount of work they were doing, such as ticket volume and one-touch tickets, rather than on quality of work.”
Seeking to make an immediate impact on customer satisfaction, Dialpad’s QA team began identifying interactions with low CSAT scores caused by lack of agent empathy. It then used MaestroQA to grade these interactions and provide helpful feedback so agents could correct course.
“We looked deeper and graded calls with MaestroQA scorecards, which really cemented that relationship between low-empathy interactions and negative CSAT scores,” Gaeta said.
Enhancing Dialpad’s QA rubric allowed the team to provide richer insights and accelerate the impact on agent performance.
“A lot of our questions were simple yes-no answers, but not every interaction is black or white when it comes to showing empathy,” Gaeta said. “We changed the scale from yes-no questions to a three-point grading scale, which led to more meaningful feedback for agents.”
Feedback from MaestroQA also provided agents with much-needed reassurance for negative interactions that were outside of their control.
“Sometimes agents have a beautiful interaction and follow the process perfectly, but that is not reflected in the actual CSAT score,” Gaeta said. “Grading interactions with MaestroQA gave agents relief for these types of situations and opened the door to more productive coaching opportunities.”
Zeroing in on problematic interactions was just the first step. Dialpad’s CX team took things to a whole new level by reimagining its definition of agent performance. Two operational metrics—number of live touches and FCR—were completely eliminated and a greater emphasis was placed on Dialpad’s quality-related metrics.
“We cut down on the number of things that we measure and increased our focus on assessing quality,” Gaeta said. “80% of what we measure is now focused on quality in terms of the agent’s QA score, responsiveness, and CSAT score.”
Dialpad experienced a near-immediate improvement in CSAT by aligning support metrics with its company values.
“The first month after we made this change, we saw an increase to 88%, which we were all very excited about,” Gaeta said. “And, we’ve steadily improved each month after that to our current CSAT score of 90%.”
Agent performance improved, too—as evidenced by Dialpad’s steadily increasing agent performance score.
"We got buy-in from agents, and we gradually started to see increases after the first two months,” Gaeta said. “Since then, we’ve gone from a 91% average score to as high as 97% agent performance.”
Dialpad’s support agents agree that the new metrics and processes make their jobs easier, too.
“Agents are using their voices, asking more questions, and really trying to learn how to make customer interactions better—instead of just closing out the most amount of tickets,” Gaeta said. “Agents love that we’re focusing on quality over quantity, and you can definitely tell there has been a huge boost in confidence across our entire support team.”
Elevated confidence and feedback from support has bettered agent performance and experience. It’s also cultivated a healthier flow of ideas that leads to even better customer experiences. Case in point, Dialpad recently implemented two new support-related systems—Guru and Lessonly—to further increase the knowledge and productivity of its agents.
Need a better way to overcome negative interactions that lead to low CSAT scores?
MaestroQA provides a scalable way to grade support interactions, provide feedback to agents, and understand the relationship between agent performance and quality.
Request a demo of MaestroQA.