Ask the Workforce Wizard

Question:

We are struggling to balance the objectives in our customer service call center. Our service level goal is 80% in 120 seconds and abandon rate goal is under 10%. Our occupancy goal is 80-85% and shrinkage goal is 25-28%. Our actual achievements are 25-50% service level and abandons of 10-20% most days. Occupancy runs 90-98% and shrinkage is 28-32%. However, we are exceeding our goal for customer satisfaction and that is what our management focuses on so they feel little urgency to address the failure to meet our other goals. How can we find the right balance and help our management understand the bigger picture?

Answer:

First of all, I need to make some assumptions. If these are incorrect, then the advice may need some revision.

  1. The reason that the SL and abandonment targets are not met is an overall shortage of staff. Part of this is a function of higher than desired shrinkage and the rest may be a budgetary constraint. However, if a significant factor is that the forecast workload (call volume X AHT) is lower than the actual consistently, that is an entirely different problem to solve.
  2. I am assuming that there are not wild fluctuations in the service level and  abandonment rate across the hours and days. If there are, then the problem may be more of staffing to the demand than a general understaffing overall.
  3. Because this is a customer  service function, and essentially a captive audience of your current customers, there is little incentive to spend more to improve the results if it doesn’t correlate with customer  satisfaction. No direct revenue is a result of these contacts.

Now for my thoughts:

  1. While some might say that 80% in 120 seconds is a goal that makes callers wait too long, there is some evidence that callers are not as disturbed by a longer wait than many centers believe. Some studies show that folks will wait up to one minute without perceiving any issue with it. And other studies show that the wait can be longer depending on the call AHT. Short calls are expected to be answered faster than long ones. Customers will tell you what their tolerance is. Generally, if they get great service and first call resolution, the wait is not so important. You generally need a goal that balances cost, customer satisfaction, and agent retention (although I have worked with some  centers that put one of these three much higher than the others in their plans and that is a business decision for each organization).
  2. The abandon rate target is subject to some considerations of what other choices a caller has to complete their business. Can they do it on the website, through the IVR, or other options fairly easily? If so, then a higher abandon rate is more acceptable. If calls are about temporary technical issues that your IT department is working on, and the queue announcement tells the callers that you know about the problem, then abandons generally indicate that the caller got the information required and is satisfied without talking to someone. That being said, if the caller cannot solve the problem any other way than to talk to someone, then an abandon rate above about 5% would generally be unacceptable as these folks may seek alternatives with another vendor.
  1. A study of customer abandon behaviors would be beneficial. This can generally be done by your ACD. The report would provide a histogram of the abandons so that you know how many leave at 5, 10, 30, 60, 120, etc. seconds of wait. Typically, there is a relatively low rate in the early queue but the curve takes a sharp upward turn at some point. That is a good sign of what caller tolerance level is and what your goal wait should be.
  2. If you have an expected wait time announcement, the tolerance for the wait is generally longer than without it. But be sure it will give a reasonable accurate forecast (which can be difficult if you use skill-based routing).
  3. When the queue is unusually long, a call back messaging system (such as Virtual  Hold) can tell callers it is long and offer them the opportunity to key in a number for a call back and hang up without losing their place in line. This can greatly extend the wait tolerance and is often substantially less costly than adding staff for peak situations. But if you have a high percentage of repeat callers or use it most of the time, tolerance for this kind of thing will decline over time.
  4. Occupancy is a result of the workload, staffing, and service level. You cannot manipulate occupancy without an opposite effect on service level – more staff to improve service lowers occupancy for example. Your current rate of 90-98% is relatively high if it is an across the board average and may contribute to absenteeism and even turnover. With a queue of three minutes, there is almost always someone waiting to be answered so your agents rarely get a breather between calls. I recommend that you do a scatter diagram to chart the impact of occupancy on shrinkage, then on absenteeism, and then on turnover. The reasonable occupancy rate will likely be revealed. But be prepared to match up whatever service level is needed to achieve that occupancy level as you cannot get one without affecting the other. When the cost of absenteeism and turnover is taken into account, a better service level and lower occupancy may be a good business decision.
  5. In addition, shrinkage of 3-4% can be very costly if it is from controllable behaviors. That is roughly a loss of 1 to 1.5 hours per person per week. At an hourly wage rate of $20, that is $1500 to $2300 per week with 80 agents. That can add up $75,000 to $100,000 per year so there is a lot at stake. Using some of this potential savings to provide incentives to lower the shrinkage can be effective. And getting that additional time back on the phones will improve the service at no additional cost. However, recognize that high occupancy rates often result in higher shrinkage as agents will find a way to get a breath somehow.
  6. If all of this is considered and management is still happy with the current performance levels, then it is time to revise the other goals to match those that achieve the CSAT result. There is no future in frustrating yourselves with  unattainable goals if management is not prepared to support them with budget. There aren’t any easy answers – just thoughtful consideration of the business tradeoffs of each option available. Reasonable and attainable goals agreed upon by all involved are important to the whole operation.

Have a tough question?

Send it to wizard@swpp.org and we’ll try to find an answer!