Elevating customer satisfaction: How I helped Wells Fargo improve sales chat CSAT by 7%

TL;DR:

In 12 months, I helped Wells Fargo's Customer Experience team elevate sales chat CSAT scores to 80% by focusing on agent experience, conducting in-depth research, and implementing targeted solutions for different agent performance cohorts. This resulted in significant improvements in both efficiency and customer satisfaction.

The challenge

When I joined Wells Fargo in 2019, our primary goal was to achieve an 80% top-box score in sales chat. Our sales agents were currently averaging 73%, and we didn’t have a clear path to enable 80% of all customer interactions to be rated five stars.

Step 1: Leading the listening & discovery process

I spent my first 60 days leading a deep dive into understanding the core problems impacting agent performance and customer satisfaction.  

Here's how I approached it:

  • Influencing stakeholders: I demonstrated to my manager and peers the value of using user experience design practices, such as qualitative and quantitative research, to help us determine what influences our CSAT scores.

  • Collaboration: By collaborating with stakeholders across different teams, such as customer success, quality control, and training, I was able to gather a diverse set of insights into what is influencing our CSAT scores negatively and positively.

  • In-depth research: The research included observing agents in real time, analyzing chat transcripts to find recurring issues, and reviewing past studies.  

Step 2: Categorizing the agents by performance

While analyzing transcripts and CSAT scores with my manager, we both noticed a trend that allowed us to categorize agents into three cohorts based on their performance:

Categorizing our agents in this way allowed us to understand each group's unique challenges and tailor solutions to help them improve. We also gathered continuous feedback on our solutions by forming an advisory board of customer success agents representing each performance cohort.

Overcoming a challenge

There was a misconception that updating the chat widget's user interface would be the solution to increasing our CSAT scores. My research quickly revealed that the root cause of low scores dragging down our average was more profound than just the UI.

To overcome this challenge, I took the following steps:

  • Presented compelling evidence: I showed examples of chat transcripts showing agents using the existing UI achieving high CSAT scores and examples where they struggled, highlighting that the UI wasn't the determining factor.

  • I developed empathy. By emphasizing the agents' perspectives and challenges, I helped stakeholders better understand the issues.

  • Focused on tangible outcomes: I linked the proposed solutions to measurable improvements in agent performance and customer satisfaction, demonstrating the potential return on investment.

By addressing this misconception head-on, I was able to redirect the focus toward solutions that would genuinely improve agent experience and CSAT scores.

Top cohort: encouraging personality

It was common for our top agents to depart from the company's scripted responses and inject their own personalities into their conversations with customers. However, this led to friction between quality control and the agent.  

  • Solution: I successfully advocated for our agents to have more flexibility in adding their personality to interactions and influenced the quality control guidelines by which we assessed agents.

  • Impact: We compared the CSAT scores of high-performing agents who injected their personality into chat conversations to those who followed the Wells Fargo brand script, and we noticed that these personality-driving agents consistently scored higher than those who followed the script. The result of demonstrating this was to discontinue flagging agents who injected their personalities rather than sticking to the script. Top agents improved their CSAT scores by 5%.

  • Key Takeaway: Personalization matters. Empowering agents with flexibility boosted both satisfaction and engagement.  









Middle cohort: simplifying workflow

Middle agents were overwhelmed with up to 14 open systems, slowing response times.  

Example of a chat agent desktop

  • Solution: I proposed a simple yet effective idea: give agents a third monitor to help them organize their systems better.  

  • Impact: Response times improved after a successful six-month pilot, saving around 111 monthly hours. These agents improved their CSAT scores by 6%

  • Key Takeaway: Sometimes, the most straightforward solutions can lead to the best improvements.

Bottom cohort: Improving communication

Bottom agents relied heavily on scripts, which hindered their ability to listen to customers and resolve issues effectively and actively.  

  • Solution: I researched and identified a technique used by the customer success team called "Huddle Guides" to help improve their agents' skills. I collaborated with the customer success managers to create targeted huddle guides to enhance the skills we identified through reviewing transcripts and finding themes around areas that consistently scored low.  

  • Impact: After providing targeted training on the differences between chat and phone communication, we measured a 24% increase in top box CSAT scores for our three lowest-performing agents. Overall, our lowest-performing cohort improved their CSAT scores by 7%.

  • Key Takeaway: Better communication skills lead to higher customer satisfaction. Active listening is essential. 

Example of huddle guide created for chat agents

The results

After 12 months, our sales chat team hit the 80% CSAT goal! The improvements weren't just numbers on a board—they directly resulted from a better agent experience, which cascaded into improved customer interactions. I also observed that the sense of ownership and empowerment we provided our customers’ success agents led to increased job satisfaction.

What I learned

Agent experience drives customer satisfaction. Understanding agent pain points is crucial to enhancing the customer experience. Data and empathy are equally important. You can't solve problems without understanding the human side of things. Simple solutions can have a considerable impact. In many cases, low-cost changes like adding a monitor or updating guidelines can drastically improve performance.

Reflection

This case study taught me the power of focusing on the people behind the product—both agents and customers. Getting granular with each cohort's challenges allowed us to implement targeted solutions. It also validated that engaged employees are likelier to stay with their companies, leading to higher job satisfaction, lower turnover rates, and improved operational performance. (1,2)

Key takeaways

  • Spend time listening to both your users and employees.

  • Segment your audience to tailor solutions more effectively.

  • Sometimes, the most straightforward solutions lead to the most significant impact.

  • Build a strong feedback loop with your team and empower them to contribute to solutions.

  • Invest in agent experience to improve retention, reduce costs, and enhance performance. (3)

Citations

  1. (n.d.). The Impact of Increased Employee Retention on Performance in a Customer Contact Center. Columbia University. https://www.columbia.edu/~ww2040/RetentionMSOM2006.pdf

  2. (n.d.). How Improving Agent Experience Can Impact Retention. Litico. https://www.lightico.com/blog/how-improving-agent-experience-can-impact-retention/

  3. (n.d.). Onboard, resign, repeat: The cost of call center agent turnover. Tethr. https://www.tethr.com/blog/cost-of-call-center-agent-turnove