Like all humans, contact center leaders are driven by the brain’s insatiable desire to experience PLEASURE and in so doing avoid PAIN! I think I have a PLAN to accomplish this mission.
Planning is essential to business in general and to contact centers in particular. While I believe that metrics and measurements are firmly entrenched in management’s mind, I am not so certain about planning. Measures have become a solitary quest, with planning taking a more reactive role. The role of planning has become one of trying to decipher exactly why the measure—the target—we were aiming for was missed. Not wanting to report any failures (or incur “penalties or pain”), we work diligently to avoid missing the mark. Hence, planning is relegated to correcting the aim and avoiding the pain of another “failure” rather than evaluating the outcome at a systemic level.
Let’s look at just two of the traditional and ever-popular measures that contact centers globally eye as evidence of effectiveness and relentlessly pursue to avoid pain and gain pleasure… abandon rate and service level.
Abandon rates occupy much of management’s frontal lobe. Desperate to “hit” a particular number, and often without a clue as to where the target number originated, management lines up time, energy and effort to “hit” this objective. Anyone in the direct marketing world knows the focus that executives put on this metric. While it is essential to manage the handling of contacts arriving in the contact center, it is also essential to determine where time, energy and effort are best positioned to produce the desired result. This is the true pleasure of confidence and achievement.
Evaluation of day-to-day abandon percentages may not deliver the highest yield in terms of energy, effort and results. The true and objective analysis is really up a couple of levels. The only real measure of effectiveness is the customer experience… whether the customer is experiencing pleasure or pain! If customers are delayed beyond their tolerance, they hang up—perhaps with the intent to try again, perhaps not. The challenge with the popular abandon metric is that it is IMPOSSIBLE to use this target for planning purposes.
Abandon is a response to a condition; it is NOT a condition in and of itself. When management views abandon as a single indicator to be “hit,” it is likely that masterful manipulation may exist to “tweak” the output until it matches the objective. This is such a common practice in contact centers that it merits comment. When calculating abandon rates, there are all manner of manipulations we enjoy. But first, let’s exclude hang-ups prior to X seconds (say 20 or 30). There is very reasonable logic here; there is little a company can do to recover these calls. They most likely represent a wrong number dialed, interruption at the caller’s end, or simply a change of heart. In my experience, the reasons for exclusion are secondary to the fact that the exclusion reduces the overall rate and brings the total closer to the target. Otherwise, why bother? What is wrong with reporting abandon percentages as they occur—X% before 20 seconds, X% before 40 seconds, etc.? This provides a bigger picture to those receiving the information and opens the door to enlightening others as to the dynamics of the contact center.
Organizing by time of day also allows correlation to other contributing dynamics, for example, increasing call volume. Those reviewing relevant reports or requesting a target must be fluent in the language of contact center metrics. Unfortunately, many are not. It is the job of the contact center leader to educate without offending. Fluency is required around many things: random call arrival, impact of special promotions and campaigns, effect of website failures on levels of contacts, impact of the organizational model, effect of small dedicated groups over pooled resources, impact of technology design, effects of forced after-call work or delay between calls, agent off-phone tasks, increase in handle time during busy times due to slow system response, ETC., ETC.!
Far too many senior-level staff receive reports on performance and have tunnel vision regarding the contributing factors. Unless you are willing and able to educate, you can be subject to evaluation by leaders that believe you simply did not answer the phone fast enough! They are not equipped to connect all the dots as to why! That is your job. If manipulation of data is the strategy you choose for hitting your target… if your desire to avoid the pain of failure is stronger than the pleasure of confidently delivering your findings… a slippery slope awaits. Energy and efforts will shift to hitting the target as opposed to developing a plan to address all the contributing factors and recommending a target that makes sense.
As professionals, we must also accept, and perhaps explain, that low abandon rates are not necessarily proof of good service. For example, if you have “hostage customers” (e.g., they cannot call anyone else bec ause you have the contract, you are the utility, or you have their money), they MUST hang on because they have no alternative. This is only one condition that contributes to why low abandons may not be an indicator of great service and why some “studies” may reflect only the output and not the input.
Open the door to whatever planning must take place to influence positively your abandon response. If you want to “fix” abandon, examine your planning process… what you do BEFORE the result… rather than manipulate the actual result.
This brings me to the next popular metric and the place where analysis and action lay. That is service level. Service level is the only “true” planning tool at your disposal to produce a plan for delivering a customer experience that matches your customer’s tolerances. Service level, like abandon, is all too often treated purely as a “measurement tool.”
Contact centers are charged with “hitting” a target number (X percent of calls answered in Y seconds—e.g., 80% of calls answered in 20 seconds). Again, we are challenged by how this particular objective was established. Far too often it is a mystery… “Our previous manager set it up,” “Our VP picked it,” and my personal favorite, “It’s an industry standard.” Well, let me debunk the “industry standard” myth. While 80/20 may be common, it is not standard. Why? Because there are no industry standards!
Anyone with a blog, a shingle, etc., can jot down a number and label it a “standard.” The truth is that there is NO governing body and NO authority to challenge any party’s claim to establishing a service level standard. While there are certainly studies, benchmarks, and surveys, nothing is truly juried. So it is best to establish your own objectives using knowledge of your own customers, your brand, your vision for marketplace differentiation, and your budget to arrive at what service level is to be at the heart of your planning process.
If service level is not central to the planning process, it becomes just another data point to manipulate. I have seen some very smart people reduced to doing some pretty silly things when it comes to “hitting” a particular service level target. Again, this is pain avoidance. This is particularly prevalent in organizations that use service level at the contractual level. We see managers set calculations in the system that are most forgiving in calculating service level. For example, using only number of calls “handled/answered” instead of number of calls “offered” invariably provides a higher number (not a better experience mind you). Rarely, if ever, does one ask exactly how the service level calculation is done. They prefer to assume that they know this.
Service level is linked in the planning process to the use of the Erlang C formula for staffing. So in order to use service level as a planning measurement, the understanding and use of Erlang C is required. (Erlang C is the staffing formula used in most contact center workforce management systems.)
If you are fortunate enough to have a workforce management system, the work is done for you. However, the WFM system is only as good as its operator. Understanding the dynamics is critical to developing an effective resource plan. The dynamics include the obvious… vetted historical data: calls offered by group, skill, etc., in half-hour intervals; talk time; and after-call work time. This data provides history and trend validation. Management must provide information regarding the future. This includes calendar events such as promotions, campaigns, system upgrades and vacations, personnel on loan, training demands, etc. Information that impacts demand, duration or capacity must be considered before producing an accurate resource plan.
If you are among the many organizations not equipped with a workforce management system, there has never been a better time to put it on your technology “wish list.” Once upon a time, the expense of these systems made them tough to justify in smaller operations. Today, workforce management systems are available on a hosted basis. As well, smaller companies have emerged that provide less complex workforce management systems to smaller contact centers. For operations choosing to use spreadsheets and an Erlang C type calculator for resource planning, PowerHouse offers a no-cost calculator on our website (www.powerhouse1.com).
Service level objectives must be based on understanding your callers’ tolerances. Study your historical data; plot service level and abandon on a graph. What does it show you? Typically, you will see a relationship. When service level goes down, abandon goes up. Correspondingly, if you improve service level, you will also enjoy an improvement in abandon. (Of course, there are circumstances in which “hostages” will hang on regardless of delay!).
To improve service level means being able to dissect its component parts to determine whether staffing, scheduling or availability is the issue. Many leap to claim, “We need more people!” Well, that is the easy way out. Before asking for more people, you must first review how the people you have were calculated (forecast), allocated (schedule), and evaluated (adherence/utilization). Evaluate what your due diligence reveals regarding performance by time of day. For example, the reason for rising handle times could be system slowdowns or uneducated agents departing in groups for breaks or lunches. This may leave you properly staffed but improperly allocated!
Agents that take breaks at will, go off to lunch in groups, or are reassigned by competing skills or groups without discussion… any of these conditions may yield an unsatisfactory service level. So before you can justify adding more people, consider conducting an in-depth audit of resource allocation.
At the same time, taking a peak at the contribution that each individual makes to each interval will shed light on utilization (aka occupancy). If you have agents that show up on time but don’t deliver while on the schedule, it is a problem. While adherence is measured with a vengeance, utilization is not. Once the team shows up, it is critical to determine their contribution; this is much greater than simply showing up. It is dangerous to think about adding additional staff before assessing your current staff’s contribution.
While I have focused primarily on inbound voice calls, total workload must be considered. All contact channels MUST have a response time and a resource plan. Non-phone work must be inventoried, evaluated and scheduled to be properly represented in the total demand (as well as any other demands taking frontline agents out of service).
Planning Creates the Basis for True Measure
All in all, PLANNING creates the basis for true MEASURE. This approach will yield PLEASURE, not just pain avoidance. Your ability to deliver on metrics is one thing. But the ability to properly plan and measure the performance of the plan provides a robust model to educate those having oversight of the contact center. It also improves the contact center’s visibility (as well as your own) as a valued asset. What a pleasure!