“I believe CX can be a department and be responsible for the plumbing and the wiring, but it also needs to be baked into every team.”
The best CX programs are as dynamic as customer relationships — they need to consider goals, various data sources, stakeholders, measurement methodologies, and of course, closing the loop. Jim Jones has over 20 years of experience leading B2B Customer Support, Customer Success, and Customer Experience teams. We were fortunate to host him last week to learn about the components that make a great CX program. Here are our three top takeaways from Jim’s session.
The What Is Nothing Without the Why
Surveys have long served an important purpose in CX programs. Survey metrics like Net Promoter Score ("NPS"), Customer Satisfaction Score ("CSAT"), and Customer Effort Score ("CES") are benchmarks that many CX professionals and beyond are familiar with, making them an attractive choice for reporting. However, Jim cautioned that these metrics are not useful without an underlying why that provides you with the context you need to understand what’s driving customer sentiment and what you can do about it. Jim emphasized that, “if you don’t know why a customer responds in a certain way, you completely miss the boat and can’t understand what you can do differently.” Jim advised the audience to insist on a mandatory why component of any NPS survey distributed to customers. Beyond simply asking why, Jim and the audience highlighted some additional ways to understand customers’ why — one is to focus questions on specific CX factors or journey stages to understand their overall CX impact. For example, suppose you need more clarity on whether your onboarding process was effective. In that case, you could ask about the key drivers that impacted that particular experience —was it your Success team, documentation, or product performance?
Another way to understand why is to evaluate customer effort. After all, how easy or difficult it is to do business with you or to realize value from your product will always be a strong determinant of relationship health. While there are many ways to measure effort, Jim advised that effort scores be considered separately from other metrics like NPS scores.
Unstructured customer feedback in support tickets and even customer interviews is a critical source of why. For example, Jim took us through an example of how he and his team raised CSAT from 62 to the mid-nineties during his tenure at Wind River, identifying specific product defects as critical drivers of low satisfaction. One of the first steps in understanding why CSAT was low was looking at anecdotal feedback. “Anecdotal evidence showed that customers were not satisfied with support. In instances where support tickets stemmed from product defects, satisfaction was especially low. We heard about this dissatisfaction from our Sales team, and we identified low satisfaction as a reliable churn predictor.” Once Jim and his team identified the low CSAT’s root cause, they ran a pilot program across a select group of customers.
“We focused on customers in the US and Canada who were above a certain spend threshold with us, who had opened related support tickets over the last few months, and who we felt we had the most leverage to improve their experience. When we worked to address the product defects that were most impactful and rolled out better training across the North American support team, CSAT improved to 85 over six months, and after another six months, reached the high-nineties.”
Following up is also critical to making surveys a worthwhile element of your CX program. If you’re going to ask customers to take the time to tell you what they think, it’s critical to acknowledge and appreciate the feedback and tell them what you’re going to do about it. “Once customers understood that we were listening, they paid attention because they wanted to tell us more,” said Jim.
Use AI To Solve for Insight Gaps
Surveys leave CX leaders with insight gaps that are challenging to fill. Rating drivers are often unclear, and dependence on responses inhibits visibility into the majority of the customer experience. “The main reason I think that AI becomes so important in CX programs is that traditional survey methods are becoming a lot less effective than they once were in all industries from technology to consumer products and manufacturing,” Jim noted. In both B2C and B2B contexts, people are oversaturated with surveys. Jim recommended ensuring that no customer gets more than one survey every 30 days, even if there are multiple interaction points across those 30 days.
Even when exercising caution about over-surveying your customers, Jim cited a 30% response rate as about the best that CX teams running email surveys can expect —and even then, you’re not hearing from over two-thirds of your customers. “When it comes to online or in-app surveys, response rates are so low that we’re almost getting to the point where it’s not worth the effort.”
Jim explained that the responses you do collect are just the tip of the iceberg and can often fall into one of two predictable buckets. “You’re getting the ends of the bell curve, the people who are very happy or very upset, and you’re not getting what you need, which is that mass in the middle. ” There are so many different channels that customers use to connect with your company, from voice, email, chat, tickets, and in-product messaging. “AI will allow you to grab that data and do real contextual analysis to give you a better understanding of what’s driving your customer experience.”
The data from voice, email, chat, ticket, and in-product messaging channels is unstructured, compared to surveys, representing structured data. Both are important, but Jim highlighted that AI’s role is to understand the unstructured data that captures your holistic customer experience. “Unstructured data is the rest of the iceberg that’s below the water.” AI can measure data points that surveys can’t, like trends driving customer sentiment. Jim also noted that unsolicited feedback from these channels complements product usage data by offering deeper insight about how customers perceive the efficacy of your product. “Natural language understanding in its simplest form is analogizing words like ‘bug,’ ‘defect,’ and ‘flaw,’ and saying, ‘these are customers that are reporting something bad about our product,” explained Jim. AI-powered CX solutions help companies aggregate feedback from all sources and measure all customers’ customer experience, whether or not they responded to a survey.
Jim also offered some advice for CX leaders who are pursuing AI-based approaches to their CX programs.
“AI will never replace the human element. In customer success, we talk about high-touch, low-touch, and tech-touch. The human element is really for your top-tier customers. Before the pandemic, I spent a week on the road and met 7 of our most important customers. That doesn’t scale, but it’s a much deeper dive. AI is for gathering customer sentiment at scale.”
Additionally, Jim emphasized the importance of targeting initial AI efforts around a specific use case rather than attempting to boil the ocean. Initial use cases could include improving or understanding CSAT fluctuations, a retrospective on why a product launch didn’t perform as expected or diagnosing top customer sentiment and effort drivers.
Tribe Structure Benefits CX Alignment
Jim highlighted an important distinction between a traditional cross-functional CX collaboration model and the more holistic model he sees as critical to a high-performing CX program today. He described the old model as Success, Support, and Sales working on their own siloed tactical initiatives, with a dedicated Customer Experience team horizontally focused on running surveys and making recommendations. Now, every department needs to engage in a unified approach to how the entire organization can improve customer experience. “While Support, Sales, and Success all still have their different functions, they operate in a flywheel with CX at the center. We put the customer at the center of everything that we do.”
Jim is enthusiastic about an agile approach to CX to reinvigorate how different teams can maximize their individual and collective contributions. “With people from different functions grouped into tribes, all aligned around a specific piece of technology that they support, it’s much easier to align people cross-functionally around what’s best for our customers.” Furthermore, Jim mentioned that product management, too, is moving toward persona-based design in agile development. “Product teams don’t write use cases. They write user stories. So if you can give your teams actionable qualitative and quantitative data to build on, that’s a home run.” This point is an excellent application of Jim’s emphasis on supporting what with why and every team’s need for quantitative and qualitative data to guide customer-centric decisions.
Thanks to Jim Jones for volunteering his deep insight and expertise and to our incredible audience for a great discussion. Check out the full recording below for more.