The contact center planning problem—developing long-term forecasts, requirements, hiring plans, shrinkage plans, extra time plans, under time plans, and budgets—is a very complex and difficult endeavor. Despite the difficulties, if your long-term plan is developed and managed well, and if it is efficient and accurate, our experience is that it will absolutely save your company money, and it will lead to a much smoother operation.
The technology available to most companies to manage this process (an Excel spreadsheet) is inadequate for the task. The limitations of Excel require analysts to take shortcuts in order to mechanically put together their plans. The complex contact center environment has outgrown the spreadsheet technology, as spreadsheets cannot accurately model multi-skill, multi-channel contact centers. The shortcuts that planners often take with their spreadsheets can be costly to the business.
The following is personally a little depressing: I am an expert on other people’s contact center planning spreadsheets. But it will allow me to say this with 99.9 percent certainty: I can look at a contact center planning spreadsheet and quickly find ways to save your company tons of cash. I’d like to share some tips that can save your company a lot of money.
- Flat-lining shrinkage is always a bad assumption
One of the easiest things to spot in a planning spreadsheet is a flat shrinkage assumption. If shrinkage is flat across all weeks of the planning spreadsheet, it means that the business expects no variation; shrinkage is assumed to be the same every week, be it mid-February or mid-July. We know this is simply not true. For instance, people call in sick in very predictable seasonal patterns.
Shrinkage is very important to get right- our experience has shown us that the shrinkage forecast is as important to correct staffing as are volume forecasts. If I miss my shrinkage assumption for one week, it will mean I am understaffed or overstaffed and it will cost us either in poor customer service, overtime dollars, or idle time.
To demonstrate this, let me ask you a silly question: Would you ever “flat line” your contact volume assumption? Of course, you wouldn’t, yet seasonal variation in shrinkage is as much as ten percent of the workforce and many of us still assume it is flat.
Examine the following graph. In this graph, we are plotting actual center shrinkage against a “flat line” assumption. It is easy to see that this assumption will cause your spreadsheet to miss plan.
Simply tracking the seasonality of each shrinkage item (e.g. by center and staff group) and forecasting these shrinkage items using the tools you currently use to determine volume forecasts is a good way to get started. One company that we work with saved 4 percent in agent time and ran a smoother operation in the process, simply by developing realistic planned shrinkage forecasts.
- Manual hiring/OT/UT plans are too hard to do well by hand
This tip is near and dear to my heart because I have a painful experience associated with developing hiring plans. Years ago, I was working at a large credit card company and was asked by our exec to do several what-ifs under different volume and handle time assumptions.
Our planning process at the time was spreadsheet-based, and our hiring plan, like most spreadsheet-based tools, involved many manual processes. In effect, we developed a weekly staffing requirement, and had to manually determine where and when to hire agents. We had seven centers, and many staff types per center.
When you think about it, this is a very complex task. We must consider- by each center and each staff type– learning curves, attrition, sick time, handle times, volumes, training plans and much more. Determining the best hiring plan, the best overtime plan, and the best controllable shrinkage plan is very difficult to do by hand looking at over/under charts.
It took a fair amount of time- maybe a week to walk through all the staff plans and my results were somewhat counterintuitive. In a nutshell, I had lower costs for a scenario that had more calls, which we all know is wrong. As an analyst, there is nothing much worse than having to tell the big boss that you need more time because your numbers are wrong. What is worse is bringing an analysis to a meeting and having the boss show you that your numbers are wrong!
Manual processes are prone to manual mistakes. Worse yet, processes that are highly complex and mathematically combinatoric like scheduling agents or determining hiring plans are difficult to do well by hand. You will always do better by developing an automatic hiring algorithm to work through all the combinations of possible plans.
This, by the way, will save your company a lot of money.
- Treating all centers as though they are the same is dangerous
One thing we contact center managers know is that all centers are not the same. While they are all similar, they each have their own behaviors and seasonality. A buddy of mine told me a great and simple example about how his company made a major mistake by moving agents and workload from one expensive center to a cheaper one by measuring cost per agent.
The company failed to take into consideration that the expensive center was much more efficient than the cheap center. This move almost doubled their cost per sale.
The lesson here is that there are all sorts of differences (such as when people call in sick, or their handle times, or their ability to sell, etcetera) across geographic centers that must be incorporated into your staff plan.
- Stretch goals are evil
One easy mistake to spot but much harder to cure, is the planner’s bane: the stretch goal.
The budget plan is, by its nature, a political document. Many managers have their hands in the assumptions and forecasts that go into the plans, but few have much responsibility for these same assumptions.
A common stretch goal scenario goes something like this:
- The planners produce a series of forecasts, including many of the major cost drivers (AHT, volumes, shrinkage, sales per call, customer satisfaction, etc…).
- At the same time, all the managers are vying to get their pet projects into the budget for the next year. As part of that process, projects are ranked by measure of benefit to the operation.
- The projects selected for funding are paid for by the improvement in one of those cost drivers, and those cost drivers in the strategic plan are altered by the amount promised by the manager.
- The benefits are not realized.
- The project timeline is extended past what is assumed in the budget.
- The benefits were not really there or were less than expected.
- The benefits may become apparent, but only much later.
- The project crashed and burned.
- Forecasters are blamed when the plan gets out of whack (note, the project sponsoring manager usually is held blameless).
- Depending on the benefit envisioned (e.g. AHT reduction), the center network may be under serious staffing pressure. Service levels are erratic.
I’d offer this advice: Don’t plan for the benefits until they are actually being realized.
- Always hitting your service goals may not be optimal
This one may sound counter-intuitive. Of course, we want to hit our service goals!
However, unless you are contractually obligated to maintain a service standard, following a service standard blindly is bound to weigh your organization down with too much cost.
As analysts, one of the simplest and most valuable exercises we can perform for our senior management is a cost versus service trade-off. Almost every decision we make has, at its core, a service and a cost repercussion. It is best for the company that we often draw out this trade-off.
Given our contact center’s seasonality, our cost structure changes as our economies of scale change and as the levers we must pull in order to reduce costs change. For instance, it is much more expensive- per call- to hire to our seasonal peak than it is to hire at other times of the year.
Many years ago, I was working for an airline who wanted to know how many people to hire, given that there was a major fare sale planned. If I drew a graph of staff versus service, my answer would be simple: it’s a lot.
However, given that the sale was so deep that our customers would be willing to wait longer to receive such a great deal, the problem became much more interesting. If we looked solely at profit only, the benefit of the marginal sale to the company wouldn’t be worth the extra time it would take to possibly secure it. Instead, the real answer was to hire only that number of agents that paid for themselves. In short, the optimal solution was to miss the standard service level, and to hire much fewer than “a lot”.
Next week, we will provide you five additional common mistakes. Stay tuned for part 2.