Will You Be Able to Deliver a Great Customer Experience in 2020?
By Eric Hagaman, Aspect Software
2020 is only four short years in the future, but the ways in which humans communicate with each other are changing at internet speed. Customers want to communicate with businesses in the same ways they communicate with each other, resulting in inescapable tectonic changes in the contact center. A recent study of high-tech enterprises by NelsonHall forecasts dramatic changes in the mix of agent-based customer communication channels by the year 2020 – changes that will put unresponsive enterprises at a significant competitive disadvantage. The chart above summarizes NelsonHall’s predictions. In it, chat usage doubles, social media usage triples, video chat usage increases by a factor of 10, while both voice and email usage decrease about 25%. That’s a recipe for a very different kind of contact center. Most of us are already using many channels that have not been fully embraced by contact centers – especially social media, SMS and videochat popularized by Skype. The NelsonHall study indicates that customer preferences and cost reduction are the primary drivers of the change, so there’s no doubt that these channels will be coming to your contact centers in the near future.
Historically, the majority of interactions between consumers and businesses have been through voice (inbound and outbound) and email channels, with minimal volumes in other agent channels. From an intra-day planning point of view, voice and email are often the perfect complements. Voice communication is characterized by long, uninterruptable exchanges (i.e., a traditional conversation) with a strong expectation that an employee will begin work on an inbound call in seconds or minutes.
Email, on the other hand, in its most commonly used form, consists of short, interruptible one way transmissions which usually carry low expectations of an immediate response. This allows emails to be serviced in the short breaks between voice conversations through universal agents, or during the longer breaks when call volumes are low through dedicated time blocks, without any customer perceptions of a poor customer experience. On rare occasions, certain types of email communication is considered to be critical, and has the same service delivery expectations as an inbound call, and work on the email is not interrupted. From a planning perspective, email has backlogs which carry over from planning period to planning period, and shares this characteristic with back office work.
At a foundational level, we can think about customer communication channels as having two primary attributes that affect the complexity of the workforce management models required to forecast and schedule for them: (1) the expected response time (aka average speed of answer) and (2) whether the conversation is interruptible.
A short expected agent response time, such as for voice calls, is the primary driver for the attention that must be given to staffing sufficient skilled agents, so contacts can be handled on average within the service quality goal. Of course, this in turn, requires that the workforce management system have associated robust forecasting models and flexible scheduling algorithms.
The ability of an ongoing conversation to be “interrupted” is a measure of whether portions of the conversation can be time shifted into less busy periods and whether other work can preempt the conversation. The quintessential example is email. It can be serviced when call volumes are low, and if call volumes spike or if a more important work task arises, work on the email can be deferred. Sophisticated modeling of email conversation dynamics is not commonly used when scheduling agents for email today, but with the introduction of additional channels of work, this will become more important.
Let’s consider a mapping of the traditional and rapidly growing channels using these two primary attributes. As we mentioned above, voice and email are at opposite extremes, and that makes it easy to combine the two in a contact center. But what about the other channels that are growing at astronomical rates? These are largely hybrids of the voice and email extremes, so they pose unique problems that traditional WFM solutions are not prepared to address. Most contact centers will not be able to manage a new set of channels at high volumes and still deliver a good customer experience at minimum cost by 2020. Consider these new channels:
- Videochat – The dynamics of a video chat are quite similar to a traditional voice conversation, in fact, videochat is just a traditional voice conversation overlaid with the video channel. Video chat is not in common usage yet, but expectations of getting a rapid answer are probably higher than for a voice channel, since the level of engagement in the conversation is likely higher than with voice alone. You are in luck, because you will be able to use existing inbound voice WFM modeling techniques to address coming volumes of videochat.
- Webchat – One of the most difficult challenges in the brave new world of customer communication is webchat. By 2020, it will be the second most popular means of communicating, but expected response times are fairly short, so it warrants attention to accurate forecasting and scheduling. Segments of chat conversations are distributed over a long enough period of time that they are somewhat interruptible, at a minimum by other chat sessions, so multi-session chat cannot be modeled with traditional approaches taken in most WFM tools. Traditional models assume that the length of the conversation is independent of the number of conversations offered (and don’t have a notion of simultaneous conversations at all), but for multi-session chat, the lengths of the conversations stretch with more chat sessions being handled simultaneously. New types of service delivery goals play a role with Webchat, and in fact may be more important than traditional goals. The level of employee responsiveness to the customer is a key new service delivery goal in chat environments. Rather than measuring only how quickly work on a chat interaction begins, we have to focus also on how responsive the employee is in returning messages. Effectively there is a service level on each message rather than just the first one. Aspect Software has done extensive research in this area and has developed new accurate multi-session chat models using sophisticated and proprietary mathematical techniques.
- Social Media – Social Media is projected to be tied for third place in popularity as a customer communication channel by 2020. At present, social media conversations take place over a long period of time, so they are interruptible, and you can’t use traditional WFM modeling techniques. They also have a much longer expected response time, so they can be planned using some of the same techniques as email such as a dedicated pool of agents or dedicated time intervals. However, a recent survey by Groove indicates that 42% of customers expect a response in less than an hour, and as social media comes into more common usage for customer care, expectations for rapid response will likely grow, and social media, like chat, will need to be modeled with mathematical accuracy.
- SMS – Texting is increasingly becoming a standard means for humans to communicate worldwide. In a survey commissioned by Aspect Software, 72% of those interviewed said they would prefer texting to picking up the phone for customer service. Like email and social media, SMS conversations can occur over a long period of time and expected response times are not immediate. However, like social media, with increasing usage as a customer service channel, consumer expectations will grow and eventually some mathematically accurate modeling will be required for this channel as well.
On top of the novel conversational dynamics of these new channels is the issue of abandonment. Two of these rapidly growing new communication channels (webchat and video chat) are subject to customers abandoning the contact attempt if they are forced to wait too long in the queue. If that is not enough, even in 2016, webchat faces issues of abandonment not only before the start of the conversation but also at any point along the way if the customer is forced to wait too long for a response during the course of the conversation. With the heightened expectations likely by 2020, social media and SMS will also be subject to abandonment as the customer switches to other more effective channels, which will appear as channel escalation – customers seeking more responsive channels. Companies not properly delivering on service quality goals for each channel will be at risk for customers abandoning less responsive channels and flooding voice channels with disruptive effects. Abandoning a conversation is a clear and unequivocal rebuke. It’s the customer changing his or her state of mind from neutral to negative with associated implications for NPS and customer loyalty. In the omni-channel world, workforce management must take a centralized view of volumes across channels and use much more sophisticated models to minimize abandons.
Our miraculous technological advances have created readily accepted new means of human communication, and consumers expect enterprises to communicate with them in the same way that consumers already communicate with their friends and relatives on their smartphones and other mobile devices. By 2020, we will have seen a fundamental transformation in the contact center’s range of agent communication channels. A new breed of WFM and other WFO tools will be required to ensure that the customer continues to enjoy a great experience made possible by more sophisticated management of the workforce.
Eric Hagaman is product manager for Aspect Software, focusing on Aspect workforce optimization technologies with a particular emphasis on workforce management. Eric monitors the pulse of the market to identify new trends and approaches to workforce optimization, looking for those product enhancements that will provide the most value to customers and help them master the next generation customer contact.