Signals: The future of VoC

0
663

Share on LinkedIn

Surya SS, Head of CXM -Tredence Inc & Praveen Satyanarayana, Director, CXM- Tredence Inc.

Over the last few years, the Voice of Customer (VoC) has become essential to customer experience (CX). Across all industries, companies use metrics such as the Net Promoter Score (NPS), Customer Effort Score (CES), and Customer Satisfaction Score (CSAT) and to measure these, they use surveys to collect feedback. Just last year, 93% of CX leaders told McKinsey that they used a survey-based metric yet only 15% were satisfied with the result, and a meager 6% were confident that this enabled both strategic and tactical decision-making.

This is the result of a fundamental challenge with surveys as a means of understanding the Voice of the Customer.

  • Surveys offer limited feedback: A minority of customers share feedback or complete surveys. These often tend to be customers who are extremely happy or irate, skewing responses to extremes. Moreover, the samples are not always representative either.
  • Sampling frameworks impact survey outcomes: Customers are likely to drop off if they see lengthy and complicated questions. When they do respond to such surveys, they end up giving similar responses to all queries — like rating 7 on all metrics. This results in poor-quality data.
  • The NPS question has inherent pitfalls: The NPS is logical for novelty items with an essence of discovery. For example, I recommended a new Thai restaurant to my colleague because I discovered it and want to take the credit for suggesting it. Why would I recommend the courier service I use every time?
  • Universal surveys miss specific target segments: Organizations tend to send a single survey to every customer once the purchase is complete, flattening the diversity of customer groups. They also do not collect insight into every step of the customer journey, making responses skewed toward the last step.
  • Surveys are reactive: Surveys are always sent after the experience is complete. Often, the feedback is received only when the customer actively notifies the organization. This can be several weeks after the experience. Additionally, surveys only describe the issue, they don’t help identify the root cause of customer sentiment. Even when this feedback is acted upon, CX corrections are made only in the next journey, which isn’t good enough for today’s impatient customer.
  • Post-facto surveys have little predictive value: NPS or satisfaction surveys cannot help predict business outcomes. At best, it measures customer experience and attitudes, not customer behavior.
  • Sweepstakes profoundly impact survey data: While experimenting with sweepstakes for an e-commerce giant, we discovered that it improved the NPS by 10 points. Customers’ perceptions might have been that promoters are more likely to win the sweepstakes, even though it is explicitly mentioned they will not.
  • Surveys don’t help in measuring the issue’s impact: With survey results, we can see that those late deliveries cause a low NPS score or customer satisfaction rating. However, surveys don’t allow us to calculate the impact of late delivery on revenue, churn, or purchase frequency. This inability to link feedback to sales leads to an uphill battle for VoC teams to persuade the rest of the organization to act.
  • To have a holistic, predictive, and precise customer experience program tied to business outcomes, VoC teams need more than just surveys. They need signals.

Signals: The future of VoC

Signals are observed feedback from following a customer’s journey through an array of captured data elements across transactional, behavioral, and operational systems. For example, let’s say a customer spends more than a few seconds on the payment confirmation page and then clicks to the support page. This behavior gives ‘signals’ that there might be a brewing problem. Just by following a customer’s digital behavior through site analytics, a potential problem can be observed and pre-empted.

Tying experience to the customer

Organizations must begin by integrating various data sources with common customer identification to track all aspects of the journey. They should consider signals across multiple customer touchpoints and build predictive intelligence on experience drivers across channels such as websites, UX studies, panel surveys, calls, audios, videos, images, etc.

Contextualizing signals, eliminating noise

Not all observations qualify as signals, most are just noise. In such cases, feedback from customer surveys, support centers, chatbots, etc., comes in handy. Contextualizing the feedback effectively into structured insights using advanced AI/ML modules. And then, use behavior and operation data of responders to identify signals within the journey that can help preempt the contextualized customer feedback.

Also, not all feedback can be translated into signals. For example, there’s no signal to preempt a customer complaint about store cleanliness, but you can capture signals for double payments/declined payments etc. As more organizations double down on their digital front door, having embedded signals make up for a missing store manager.

Focusing on the impact of the issue

Traditional VoC solutions optimize for the volume of feedback without quantifying its impact. Signals enable the extrapolation of the observations to all customers using historical data and measure the magnitude of the impact. With this, you can develop a prioritized roadmap with functional teams to resolve CX friction.

Embedding signals into your operations

Embedding signals in an operational system will actively look for customers facing undesirable experiences and establish a touchpoint in real-time. Currently, VoC platforms close the loop with responders, while signals can do the same with all customers. Orchestration tech stacks are also primed to implement such use cases and make real-time engagement feasible.

For example, a curbside store pickup operation was consistently getting negative feedback about delays. Operational metrics showed that all orders were ready before their slot times and picked up on schedule. Connecting feedback with signals showed that most of the complaining customers arrived ahead of schedule and expected orders to be ready. Based on this, the business sent notifications to early birds letting them know that they’d arrived early.

Embracing customer centricity

Customers today are diverse, and their purpose determines the type of behavioral segment they belong to. A signal driven CX program tries to understand the needs, wants, and challenges of all customer groups, providing every customer with a meaningful and enjoyable experience.

Proactively resolving issues

With survey based VoC programs, organizations wait for customers to acknowledge their challenges. On the other hand, signals turn an organization into a listening one. This means picking up on signals while the experience is happening and designing interventions accordingly.

Building the CX of tomorrow

Signal-driven CX needs a robust data science and analytics engine that combines all this data and presents actionable, granular, meaningful insights and recommendations in real-time. Here’s how organizations can build that.

Be deliberate about signals: Not all observations qualify as signals. A good analytics engine must gather clean signals data and combine it with other sources, such as surveys, and chatbots, for contextual insights.

Redesign surveys for better outcomes: Most surveys are optimized for ease of analysis. For instance, surveys will invite customers to rate the organization on a scale because it is easy to aggregate this and build a dashboard. However, this is limiting. Open-ended questions are a great way to understand what is essential to customers. They can provide in-depth insights that operational metrics/enterprise metrics fail to provide.

Setup real-time recommendations: The insights from signals can be leveraged in two specific ways. The first is to build dynamic customer experiences across touchpoints based on real-time feedback. For instance, if a customer moves from a payment page to the support page, a chatbot can be activated to ask, “can I answer a question for you?” accelerating the journey to addressing the customer issue.

The second is to regularly alert CX teams about ineffective experiences. Insights and recommendations can be sent as alerts/notifications to CX teams, including frontline agents, in real-time, to empower them to anticipate issues and educate them about potential solutions.

Measure the right outcomes. Some of the metrics organizations must consider are:

  • Conversion Rates: Customers who moved to the next stage of the journey across touchpoints, expressed as a percentage
  • Repeat Rates: A customer’s intention to return to the organization and engage, expressed as a percentage
  • Customer Lifetime Value: Average customer revenue generated throughout their relationship with the organization

When specific issues are identified:

  • Customers Impacted: Customers who faced an issue out of the total number of customers represented as a percentage
  • Severity Index: Difference in ideal repeat rate and the repeat rate due to the pain, higher the difference more painful the experience
  • Average Resolution Time: Time taken to resolve an issue once identified — the best case is to resolve it even before the customer experiences it

Customer experience is the single most important competitive advantage for any organization today. Monitoring, optimizing, and constantly improving CX has become a C-level imperative. To stay ahead, organizations need predictive approaches to gain a contextual understanding of where the experience is lacking.

Businesses need a more granular and empathetic understanding of the customer and behavioral segmentation that goes beyond demographics, psychographics, and the like. They need to anticipate a customer’s emotional reaction throughout their journey and design the experience accordingly. This needs to be done in real-time to get ahead of potential problems. All this is only possible with signals, which pave the path for the promised land of VoC programs.

Praveen Satyanaryana
Praveen is a Director of Retail Customer Experience Management practice at Tredence. He graduated from Duke University with over 10+ years of consulting experience focused on creating end to end retail customer experience & analytics solutions. He has consulted for 15+ retailers across Food, Apparel, Beauty, & general merchandising stores and helped them deliver personalized solutions using data & AI. He strives towards a vision of creating customer centric cultures for his clients & helping them compete & differentiate in these red oceans.

ADD YOUR COMMENT

Please use comments to add value to the discussion. We will not publish brief comments like "good post" or comments that mainly promote links. Maximum one educational link per comment.

Please enter your comment!
Please enter your name here