Designing a Digital Contact Center Customer Journey That Ensures Satisfaction

A digital-first contact center approach, with methodical escalation to agents for human connection and managers for resolving the most difficult requests, is key to ensuring customer satisfaction, explains Chris Small, VP product management, Roadside.

The goal of every contact center team is to resolve customer requests as quickly and effectively as possible. Sometimes they do so by moving the customer journey through digital channels, and other times they escalate requests to specialized contact center agents or management for further assistance. But it isn’t always easy knowing which approach is best for a given situation. For customers, too much time spent in digital channels that don’t resolve the issue can feel impersonal or ineffective.           At the same time, routing customers to agents or management for simple tasks that automation could support can be a misuse of precious time that could otherwise be spent on more complex cases.

How can you balance humanity with efficiency and still methodically escalate requests? The goal should be to design a digital-first contact center customer journey that inserts agents where there are gaps or opportunities to drive more human connection and leverage an agent’s high-trained skillset. As our chief digital officer, Bernie Gracy, says, “We are moving FROM business processes managed by people and assisted by technology TO business processes managed by technology and assisted by humans as an exception or escalation.”

Before we get into how and when to escalate customer requests, let’s first review some common data elements in a contact center and how each impacts customer satisfaction.

Learn More: How Personalized Live Web Chat Can Boost CX

Contact Centers Are Drowning in Data

On average, contact center agents handle up to 50 calls per dayOpens a new window . In a center with 600 agents working each day, for example, that amounts to 30,000 calls per day and 210,000 per week. With such a high volume of calls, contact centers have access to invaluable data such as when a customer called, the frequency of calls, the reason for their call and more. Each of these data elements tells the story of a customer’s journey for help or support, allowing you to closely monitor the situation and make decisions about how to best handle requests.

However, today’s technological advancements have made monitoring the customer journey and responding reactively baseline requirements; organizations today need to take their approach to the next level. A more advanced strategy involves leveraging machine learning to predict what the customer will need next to achieve a positive outcome — or an optimal level of satisfaction (e.g., a high net promoter score) — and proactively and automatically trigger agent involvement. Because of this high volume of call data collected, machine learning models offer a high degree of confidence in predicting if something will go wrong or a customer will want to speak to an agent.

Data Helps Predict Customer Satisfaction

There is a wide range of data inputs that feed into machine learning models. For instance, with 30,000 roadside assistance dispatches running through our system each day, we look at the time of day, day of the week, month, if it’s a holiday, the customer’s location (side of the road, a parking garage, their driveway, etc.), zip code, and the type of service they need (tire change, flatbed tow, wheel lift tow, etc.). As time progresses, we layer in other data elements, such as the estimated time until a tow truck arrives, the service provider selected, how much time has elapsed since dispatch, and so on.

Other helpful data components can include case notes or comments in the customer relationship management (CRM) system. Advanced machine learning models can not only scrub words and groupings of words from these notes but can also detect the inflection, tone, and sentiment of customer voice recordings.

Of all the data elements, though, the time elapsed can arguably be the most important factor in the outcome of a customer’s journey. In fact, nearly 60% of customersOpens a new window feel that long hold and wait times are the most frustrating parts of a service experience. Rather than waiting for that time to elapse, machine learning can predict the time it’s going to take to fulfill a customer’s request.

In addition, digital communication channels, such as mobile apps, on-demand apps, and online web portals can tighten operations and drive efficiencies by helping customers access self-service support more. These channels essentially help customers “cut the line,” rather than waiting in a phone queue.

Escalating Customer Requests to Agents

While the easiest, most straightforward requests can be handled through digital communication channels, more difficult requests require agent intervention. When a machine learning model detects input variables that are likely predictors of a poor outcome, it can alert agents that those requests need a heightened level of visibility.

It’s also important to note that while some customer requests may require an agent because they are more challenging issues, others are simply situational. For example, in the roadside assistance industry, a customer stranded on the side of the highway at two in the morning when it’s dark and raining might prefer the comfort of knowing they have someone on the phone giving them assurance, even if they have quick, digital channels at their fingertips. This is why it’s important to have multiple communication options available. You should guide customers into a digital channel when possible, but don’t force them to stay there if they require human assistance. It’s all about balancing efficient digital channels with that more personal connection.

Escalating Customer Requests to Management

While agents are often considered the first line of intervention, there may be situations where they’ve exhausted all the tools at their disposal and require management intervention. Management escalation typically occurs during moments of heightened emotion – think calls into an airline about a canceled flight while far away from home or during the holiday gift-buying rush when that long-awaited present arrives damaged. For roadside assistance, this is usually when a customer is in unfamiliar surroundings far from home. These extra challenging events require a delicate balance of art and science, which a management team is well-equipped to handle.

Management escalation can also give more difficult requests higher visibility within your organization. The executive team can review these requests, as well as the aggregate data logged into the CRM throughout the escalation process, to evaluate issues and opportunities and make broader business process improvements.

 Learn More: Why Intelligent Automation, And Not AI Drives Contact Center Efficiency

Designing Contact Center Journeys for Customer Satisfaction

Providing customers with quick and helpful support is a crucial part of every business. Throughout each stage of the contact center customer journey, there are time-bound milestones that a customer should reach to ensure a positive outcome. If you can leverage data to anticipate those inflection points and proactively act on them, you will be one step closer to delighting customers and building loyalty.