Historically, organizations often viewed customer care as a safety net for things that could go wrong. Marketers, product owners, and other areas of the business did the best they could to design campaigns or solutions. But if they fell short, they knew the contact center would handle the fallout.
Today, companies understand that overall customer experience is critical for loyalty and propensity to spend, and therefore the contact center should be the last place customers end up for service. They understand that engagement with their company and agents (should that even be necessary) should be effortless or frictionless, advancing the idea of Net Promoter Score (NPS) further to where the question becomes "Was it easy to do what you needed to do today?" instead of simply "Would you recommend us?" This is the true predictor of loyalty and customer spend. Especially as customer expectations are changing faster than ever, and the technologies that enable them advance even faster, leaders know that getting customer service right is essential for survival and growth.
If our objective is to reduce or, better yet, eliminate customer effort, then businesses should have a shareholder view of customer service and not a safety net view with service teams shouldering a larger proportion of responsibility. This means every contact into the contact center should be critically evaluated on how the overall organization could have done something differently to prevent the contact in the first place. Products and services should be designed with the idea that customers can serve themselves or should not have to contact service centers at all. In fact, let me go one step further and suggest that many service organizations treat customer contact as a defect. Naturally I feel a bit differently about inbound sales organizations.
With this in mind, consider these three parts of the customer service journey and the technologies and processes needed to get there.
Stage 1: Prevent the contact (top of the funnel).
In an ideal state, you have designed your product or service perfectly, marketed and delivered it flawlessly using customer data to anticipate needs and preferences, and there's likely never a reason for customers to contact you.
This should be the top priority of the customer service operation: Work with the rest of the organization to eliminate all need for contact. This also means that when the inevitable contact does occur, there are standard processes and procedures to examine why the contact occurred and an enterprise-wide view on how to fix it. Are shipment delays causing WISMO (where is my order) calls? Are bills impossible to interpret? Is access to your platform causing password reset requests? These contact types do not originate in the contact center. These are enterprise issues, not customer service issues.
There are numerous tools and processes to help organizations identify top call drivers. A pareto chart of call types intersected with other parameters, such as length of call, complexity of the call type, and availability of data to aid in resolution, will allow the company to prioritize the call types to eliminate first. The assessment can be done in many ways. Everything from reporting out agent disposition codes to doing voice and text analytics on a large data set will produce the contact types as well as the long-tail intents that are the follow-up and defining questions. These analytics activities can also be used in real time to produce reports that give insights into the nature of contacts into the center.
Stage 2: Automate the contact (middle of the funnel).
If customers do need support and end up contacting customer care, then provide them with a frictionless and effortless way to resolve the issue themselves, perhaps through a virtual assistant. While not all issues should be handled by a machine— irst Notice of Loss (FNOL) on a life insurance policy, for example, shouldn't be done by a non-empathetic machine—most issues should not require a human.
Many organizations still believe that customers service is the one opportunity to wow customers and retain them. However, it is far better to let customers instantly solve issues themselves if they can't be prevented in the first place.
For example, can the customer inquire on the WISMO request and get information that is not evident on the shipping status notice? If they want to return the item even before the package arrives or maybe try to cancel the shipment altogether, is there a way they can do that? These are answers that typically come from customer service representatives either via chat or phone, but since the answers are standard and tied to policy, they can easily be handled by a bot.
Stage 3: Handle the contact with a human agent (bottom of the funnel).
If the contact was not prevented earlier (Stage 1) and customers cannot solve the problem themselves (Stage 2), businesses need to handle contacts as efficiently as possible, ideally with artificial intelligence-empowered agents working alongside humans. Different from traditional contact center optimizations that use automation, using AI and other technologies to analyze chats and voice calls can help guide agents to the best answers. More ideally, AI would learn, prioritize, and reinforce the best responses from the best agents and propose them to all agents.
This third stage is also the key feedback loop for the entire operation. It's important to consider how data about why people are contacting the company and which challenges agents have in servicing customers gets back to the top of the funnel. In other words, how do improvements get made to the front end to remove the defects from the overall process?
The WISMO calls described above are a prime example of the feedback loop. The shipping and logistics department has a service-level agreement (SLA) to get products out the door. Theoretically, many of the WISMO contacts are likely because that SLA was not met. Sharing data from the contact center and creating a cost-to-the-company posture around the financial impact of each one of those WISMO calls will put a value to the organization for missing the SLA. Using the old, "that which gets measured gets done" adage, the organization should measure the SLA and the WISMO contacts and actively report on and manage this performance. This is one of a hundred potential feedback loops in the organization that can be quantified using analytics in general, speech analytics specifically, and better use of data in customer service.
The bottom line is that customer service is a team sport, and the goal in the end is to make it effortless for the customer.
Tom Lewis is managing director of Accenture Applied Intelligence.