Expert panel offers insights on the prevailing technology and its usage.
In spite of the fact that nearly every “call center” is now a “contact center,” we observe little substance to the discussion—much less action—around multichannel workforce management (WFM). With “omnichannel” one of the most frequently heard buzz words in our industry, we decided it was time to look at planning, forecasting, staffing and scheduling across media, with all its nuances and challenges. We engaged our own team as well as an expanded set of experts (see the “Contributors” box) to explore both the prevailing technology as well as if and how centers are using it. We started with a set of questions, and our esteemed panel responded with their insights and experience. We’ve distilled the input into our own little “state of the union” on WFM in this omnichannel world.
State of the Omnichannel Center
Contact center leaders “get it.” They want to move beyond Excel worksheets to common reporting, forecasting and scheduling that address all media needs. They want to proactively manage workload across channels. But omnichannel is not all it’s cracked up to be in the marketing hype—operationally or technologically. The issues across the people, process, technology spectrum are interrelated, creating a bit of a catch-22 or self-fulfilling prophecy.
Centers don’t do multichannel right because it’s too hard (partly because of lack of tools), and they don’t put technology in place because they don’t “blend” staff. Daryl Gonos of WFMSG highlights the “disconnect between end-users and technology providers. End-users are struggling to meet the basics and are focused on the fundamentals; tech providers are focused on moving forward to the higher vision.” But Interactive Intelligence (ININ) accurately describes contact center leaders’ reactions to discussions about the possibilities, “All in one system? All in one interface? That would be such a relief…”
Routing and Reporting
All seem to agree the industry is early in its transition on multimedia routing and reporting. People focus on voice as the historically and still dominant media. Many have a separate chat system, often cloud-based, and use manual email with Outlook. Some centers use CRM for one or both of these channels. Others just treat non-voice contacts as “project work” to handle in bunches. This “patchwork of systems and teams” (as ININ calls it) is compounded by other applications and groups for social media and back office. And if an outsourcer is involved in handling other channels, they are inherently separate.
Not all routing engines are fully integrated across media, perhaps further slowing adoption. The larger centers with bigger budgets are more likely to be the early adopters of the technology and the move to omnichannel. Those that do so reap the benefits but recognize the need to change more than technology. As Genesys says, “Businesses are trying. They want to do it… making the change to omnichannel routing is difficult. It requires changes to their business processes and strategies.”
Blending is an atypical practice and faces many challenges with different media types and their WFM. The consensus views are: (1) agents that are cross-media usually do so in dedicated blocks of time, especially for non-real-time work such as email and back office; and (2) different media require different skill sets. Multichannel contact handling and management hasn’t worked well in terms of productivity or accuracy, but part of the problem is the lack of structured WFM in these situations. Yes, voice is a different skill set than written communications. Multiple chat sessions may be feasible, but we better account for the concurrency and equip people to minimize risks. And “occupancy” as we’ve known it can take on a whole new meaning in this new world.
A growing number of studies reported on by CNN and NPR show humans are not able to multitask (even the millennials!), reflected in losses due to “switching gears” and the time it takes to re-engage when returning to another media. CNN states, “Only 2% of the population are “super multitaskers” or able to do several different activities without losing efficiency or quality.” This finding must be true for agents as much as anyone else. So blending impacts bandwidth, performance and quality. Genesys highlights how the “human factors bottleneck” also varies by person (skills, experience or newness to center, etc.), adding another variable to scheduling. As Aspect says, “The so-called ‘universal agent’ … is a rare and small group, easily overworked by excessive context switching in a true universal blended environment.”
Many express concerns on a different form of blending, weaving in non-real-time activities to fill available time. This practice can create utilization levels that lead to burnout and contribute to high turnover. Too much interruption drives inefficiency, errors and frustration as the “gear-switching” problem manifests in a different way. Further, it not always proper to have email or other deferrable work lower priority.
While voice has been the predominant volume, expectations are it will change. Other media volumes are growing in total and as a percentage, and the diversity continues to grow, especially as mobile devices play an increasingly dominant role in our communication lives. It’s hard to dispute that these dynamics would change more if centers put as much attention on other channels as they do voice!
So if blending is out, what is in? Omnichannel is not about everyone taking every media; it’s about managing all media effectively and creating a “seamless, integrated” customer experience as they use their channels of choice. Common routing, reporting and management, across channels (aka, Multimedia Routing and Reporting, or MMRR), is critical to that vision and enabling centers to create and use accurate data on all media. (See Figure 1.)
State of the Art: Routing Engines and Data Sources
To do omnichannel WFM, you need multichannel data inputs. Two things account for workload—volume and handle time (or the equivalent in other media). WFM analysts need a fair amount of detail on these things—by interval, across time and with accuracy.
Vendors long ago solved the challenge of how to get data on voice contacts from many different ACD systems into many different WFM systems. They are now solving that for other media, with a wide range of sources, many of which provide not enough, or not good enough, data. That is both a user issue and a technology issue. As ININ says, other media too often has been the “invisible workload” as centers at best use a work code to capture time spent on other media. Aspect expresses a similar sentiment that too many try to treat real-time contacts as “project work” (like they do email), but things like chat need fast response. In addition, the WFM vendors have to tackle issues with immature solutions that don’t provide adequate and proper data and account for the differences between media types and concurrency situations. Once they understand what data is available, they need to build the integration to obtain it.
Two scenarios define the landscape:
- Those with full routing and workforce optimization (WFO) create and use their own data (e.g., Aspect, ININ, Genesys), taking advantage of the value of a full suite across core contact center technology (routing and reporting) and WFO.
- Those depending on others to create and provide the data—whether a WFO WFM-only vendor or a suite that is not controlling all routing (e.g., dealing with a standalone chat or text SMS solution or a CRM solution).
Here are a few examples from the vendors of how they tackle the challenges:
ININ uses “Focus” to accumulate direct handle time (talk and wrap), while “pause” time also accumulates. Visual indicators help the agents but clearly they play a big role in getting it right, clicking on the right interaction to create proper tracking.
Verint uses the data from the ACD if they get it (and it’s accurate), but have other options. In some cases, chat systems provide the data on handle time and number of chats but must provide some information on concurrency or data adjustments will be required. In a pinch, time and motion studies may help. But a better scenario is to use core systems data or Desktop and Process Analytics to get data from systems (e.g., for back office or for time spent in applications).
Genesys outlines three typical ways to get the needed data: from a system directly, enter manually if available, or estimate when not available. Clearly these are in descending order of ease and accuracy, but the good news is they see a distinct upward trend in getting good data from the multimedia routing engine.
Aspect knows that “universal input” is the goal, as they’ve achieved with voice, but that is a work in progress for other media. They can do manual data entry for proof of concept and then integrate with APIs to get specific data needed.
State of the Art: WFM for Other Media
So once they get some data, WFM solutions need to forecast, staff, schedule and report on the variety of media. The vendors have adapted and developed solutions to deal with the different characteristics of the various media, as well as customer expectations and the resulting contact center management implications.
Service level (SL) and average handle time (AHT) and its components (talk time and wrap time) are still table stakes and critical to WFM, even as people try to focus on more strategic metrics. As Genesys indicates, “Right now speed and response are still the primary drivers… But there is pressure to move from speed toward happiness/satisfaction.” Increased focus on Customer Sat and other metrics will continue to shift the conversation, even though it may be quite difficult to factor “customer experience” or “happiness” into workforce planning! Changes to routing strategies based on personalization, customer match or other “context” can disrupt WFM further.
But for now, the focus is on what differs about these foundational elements of WFM for other media. For example, outbound contacts focus on connect rate and right party connects, while email considers response time. Chat requires SL as a real-time media, but since it can have multiple interactions in one “conversation,” it requires fast response on subsequent messages, too. Aspect frames it this way: “Each channel has its own set of interaction dynamics, timing characteristics and associated service levels,” and points out the different characteristics of media in two dimensions: response goal (fast or slow) and single continuous communication versus multiple separate messages. “Handle time is a voice notion. Other channels require different notions… each of these channels needs to be uniquely modeled and then planned concurrently…”
So the vendors must use different algorithms for different work types. Verint describes three types of communication: immediate, deferred and outbound phone, but notes the need for different algorithms for each media (“heuristic search and simulation techniques layered with channel-specific characteristics”). For example, chat is immediate (like phone), but its algorithm allows for multiple sessions. Email uses a deferred algorithm that addresses backlog as well as the longer service goal. Genesys talks about “channel-specific algorithms” and how they “simulate many different strategic staffing models and scenarios.” Aspect uses a “set of parameters to mathematically describe the characteristics of each channel’s interactions” and then runs thousands of scenarios using a Monte Carlo simulation, an approach used to model complex, dynamic systems. All of the vendors also emphasize the importance of channel prioritization and its impact on WFM.
What State Are You In?
Nearly every center is multichannel today, with more on their “wish list.” And most know they need to deliver better service on the non-voice channels, giving them the respect they deserve. Technology is the great enabler of this vision, but must go hand-in-hand with strategic operational change. Even without blending, there is value in good routing, data creation and reporting across media. That data feeds into WFM, and smart multichannel tools can create proper forecasts, staff projections and schedules, along with real-time management, for email, chat, text/SMS and more. The fundamental tool and operational principle that has been so well developed for voice contacts can now be applied to each media and all its nuances. The technology is ready. Are you?
As we look at the state of WFM, it is worth highlighting a few other areas that go “beyond phone calls.” Here are some innovations and applications of advanced WFM capabilities getting traction today:
The main WFO vendors (including Verint, Genesys and Aspect) see significant activity in managing back-office work. This area is ripe for better tracking, scheduling and optimization, but not identical to the contact center. As Aspect highlights: “In the back office, it is typically more important to precisely manage the work than precisely manage the people.” The back office in many cases ties into interactions with the front office and must accommodate differences such as variability in handle times. Some vendors, such as Genesys, offer a combination of the routing engine for back-office tasks and WFM to handle it: “We see more interest in this as omnichannel in the back office evolves, and employee engagement and customer satisfaction become increasingly important.”
We all love our mobile phones, so providing schedules and interfaces to request and make changes here, in addition to the Web, only makes sense. Some, like Aspect’s “Mila” (a natural language chat bot) also include voice-driven personal assistants. The mobile app can also tie into scorecards of KPIs for suite solutions such as Verint.
Integration with LMS and other WFO tools
Our previous two columns highlighted scenarios that show the value of a suite of tools and integration. For example, Verint, Genesys and Aspect bundle QM, scorecards, coaching sessions and learning (whether CC-specific or the corporate LMS). Aspect points out how a fully integrated tool works both ways, defining when to schedule training and to update skills based on training.
Thanks to Our Contributors!
Eric Hagaman, Senior Product Manager and Bob Moore, Senior Product Marketing Manager, Aspect
Yvonne Ba, Product and Solutions Marketing, Genesys
Chip Funk, Contact Center Consulting Manager,
Karen Torf, Senior Product Manager, and Rebecca Gibson, Senior Product Manager, Interactive Intelligence (ININ)
Jay Minnucci, President, Service Agility
Rajeev Venkat, Senior Director, Solutions Marketing, Verint
Daryl Gonos, Principal, Sales & Marketing, WorkForce Management Software Group (WFMSG)