How Digital Channels Are Changing the Contact Center

By Nick Martin, NICE

Adoption of digital channels, from email and chat to SMS and social media, has skyrocketed due to the pandemic, placing new demands on contact centers. The old ways of collecting data, calculating staffing needs, and generating schedules – then doing it all again as conditions change – no longer suffice when dealing with digital channels. In a marked contrast to the past, work today doesn’t necessarily flow sequentially, in a continuous stream of work handled by a single employee. It’s now nonsequential and noncontiguous in nature, and that has a significant impact on everything from data integrity and acquisition to staff requirement calculation, schedule optimization, and change management. 

Here’s an overview of the changes – and the workforce management challenges you need to consider as a result. 

Immediate Response & Deferred Response Contacts

Employees today are tasked with handling both immediate response contacts, such as phone and chat, and deferred response contacts, like email. How you report service objectives for deferred work, however, will differ from how you report service level for immediate response contacts, and this has an impact on historical data reporting. You also need to calculate accurate and trustworthy staff requirements for each of those two types of contacts and using Erlang to account for multiskill efficiency or taking a workload approach when dealing with deferred response contacts can be complicated in contact centers that use omni-session handling or have multi-concurrency or multiple simultaneous interactions. To optimize schedules with multi-skilled employees, you need to understand how their work is going to be distributed against the concurrent, noncontiguous contact streams.  And you need to manage your backlog and reforecast as the day progresses and supply fails to match up with demand.

Deferred Work Forecasts Based On Capacity

With a digital channel such as email, contacts are often deferred. You may, for example, give agents 24 hours to work on an email, and that has implications for forecasting. And when staff is shared, with inbound work like chats or phone calls, it can affect how work is propagated into the future. Deferred work forecasts based on capacity create new challenges in how you plan for work that is going to be spread over the course of time, considering staff capacity per interval rather than a flatline distribution of work across time. They affect how you optimize and allocate multi-skilled employees’ time to each work stream. And they have an impact on how you manage the backlog, make intraday reforecasts, and analyze supply capacity changes to adjust to the peaks and valleys of the demands you’re placing on employees.

Session Concurrency Within and Across Channels

When you have multiple interactions happening simultaneously or interactions that overlap, understanding average handle time (AHT) and utilization becomes much more complex. Your workforce management system also needs to be able to understand how each employee copes when handling multiple interactions concurrently – what one agent can accommodate with ease can overwhelm another.  How do you interpret AHT (in focus vs. elapsed) and handle intra-interaction utilizations and messaging (the amount of downtime or wait time for both the customer and the employee)? How do you interpret multi-session handling limits? Your definitions, as well as your objectives for maximum utilization, speed of answer and speed of response, will affect your staffing requirements calculations and schedule optimization. 

Channel Interrupt Priorities

One digital channel is often more important than another when it comes to delivering the service your customers demand, and many organizations prioritize work streams accordingly. Often, it’s the immediate response channels (e.g., phone and chat) that are given more attention, more quickly, by interrupting agents’ work on lower priority interaction channels. Which interactions can be interrupted, and how do interruptions affect the AHT for different types of interactions?  This, in turn, will affect how you calculate staffing needs and schedule employees. 

Long Asynchronous Interactions

While customers expect an immediate response with some types of digital interactions, others are often characterized by long gaps in time; a conversation that starts on one day can actually finish the next day or even further into the future. Do you use a hybrid approach, hybrid transaction counting, or work time concatenation to account for the gaps within a customer interaction? How you define AHT and treat the data coming into your WFM system has significant implications for staffing, scheduling, and planning.

Elevated Interactions

Elevated interactions are closely related to long asynchronous interactions in that they raise the complexity involved in calculating AHT. Consider the case of an interaction that starts as a chat before the employee elevates it to a phone call and then ultimately to email. The agent may move along with the customer, continuing the conversation on each new channel, or send the customer to the general queue for assistance, depending on your processes and policies.  Do you track average call duration for the entire series of interactions, combine phone and chat for two contacts, or break it up into three separate contacts?  How you concatenate and interpret time will affect how you calculate staffing, optimize schedules, and manage intraday and supply capacity change. 

Employee Cognitive Load Limits

The demands of contact switching due to interruptions and simultaneous interactions on digital channels add to employees’ cognitive load, or individual employees’ abilities to juggle contacts and responsibilities: overload can lead to frustration and poor decision-making. Cognitive load encompasses load per type of contact; maximum capacity per employee (overall total and per type of contact); and algorithms that ensure that “low” loads do not take precedent over “high” loads, that dynamically define load per type of contact and that dynamically adjust employee capacity. Cognitive load limits create new challenges in how you set limits across channels, track employee load-based performance data, adjust base staffing requirements, and optimize schedules, making cognitive load limit adjustments to individual employee contribution. 

Dedicated Task Time Limits

A lot of different things come into play when you start talking about task management – for example, when an employee is asked to spend time focused solely on managing a social channel. How do you manage the appropriate blocks of time, the amount of time, where it’s placed, and who gets it? And how do you do so fairly?  Dedicated task time limits create new challenges in how you optimize maximum and minimum task time constraints and ensure fairness.

AHT Longer Than Stat Interval

Work time, such as AHT, sometimes exceeds your planning interval. It could be as simple as a 15-minute interval and a contact expected to take 30 minutes to resolve; you can’t simply multiply the one contact by the AHT to arrive at the workload, because two people working simultaneously in that 15-minute increment can’t handle the same contact and get it done more quickly – work must be carried over to the next interval. You need to establish a new paradigm for carryover work, then optimize schedules for employees engaged in work that stretches across intervals while keeping track of who is engaged, and who isn’t, with work from prior intervals.  

Employee Self-Select Work Items

When employees are allowed to choose the type of contacts they will work on during their shifts, effective planning requires artificial intelligence (AI) able to interpret and predict employee behavior, so you can understand and anticipate which work items a given employee will choose and use it to generate schedules accordingly. 

Digital channels offer significant potential to better (and often more cost-effectively) serve customers needs. By understanding how common workforce processes are affected by the unique characteristics of digital channels, you can position your organization to meet the needs of your customers and break new ground on the digital frontier. Learn more about NICE WFM with machine learning 2.0 for digital channel management at 

Nick Martin is Product Manager at NICE.  He may be reached at