Bad Decision Avoidance
By Ric Kosiba, Ph.D., Vice President, Interactive Intelligence’s Decisions Group
Biggest Source of Benefit
Those of you who have read this column before know that I am a big fan of planning, I love mathematical modeling, and I have a thing for cool analyses. So I get asked, fairly often, about the hard benefits of applying these technologies to a contact center.
There are many. Better capacity planning leads to lower costs, as just-in-time algorithms will develop very efficient staff plans. Developing accurate simulation models of the contact center network will yield more accurate staffing requirements, and almost always at lower staffing levels than the Erlang equations. Better forecasting technologies can develop tighter forecasts and lead to lower variability. Actively managing the long-term plan closely will lead to more consistent service delivery.
All of these benefits are impactful. But what is the biggest benefit? By far, it’s bad decision avoidance.
What are the worst kind of mistakes executives can make in a contact center operation? I expect that we would all agree that strategic mistakes—bad decisions that can negatively affect the operation for a long time—are the worst. It is not to say that day-of service hiccups can’t ruin a day; these one-off terrible days are truly… terrible. But chronic understaffing can produce long-lasting lapses in customer service that tarnish your brand, and create both customer and agent disloyalty.
There have been many examples of these sorts of mistakes in the news: companies deciding to offshore important customer functions too rashly, operations being unprepared for regulatory changes that drive up handle times, and companies unthinkingly cutting costs and, hence, cutting service.
But there are other, more common strategies that many contact center operations have that create self-inflicted service meltdowns or issues. Let’s discuss some of these seemingly innocuous practices.
Catching Up by Overstaffing
I love this one, because it seems so benign, is such a human strategy to pursue, and I hear it fairly often. It is this: early in a month or week or even a day, there is a time period where service is significantly below standard. Because the contact center executives are judged based upon service attainment at the end of the month or week, they decide to “catch up,” to overstaff until their average service levels are at their goals.
So what happens? For the rest of the week or month, the company incurs a fair amount of overtime, the agents are both overworked and yet bored, because their occupancy is fairly low. Customers are answered with very high service levels, but being answered faster than service goals will likely mean the customers never even note it as a benefit.
Is this a service failure? Heck yes it is. Agents work longer hours and do little. Customers who were inconvenienced by the poor service at the beginning of the month see no change, since they are unlikely to call back. Customers who call when the center is overstaffed notice little other than there were few rings when they called; in other words, they will discern nothing. The center incurs significant staffing costs for no real purpose—except to make a metric look better.
High costs and a sloppy organization. What’s not to like?
The gist is this—with the appropriate analysis, the folly of this policy is made clear. More on that in a bit.
We all love to note how contact centers are the embodiment of change, and I believe that it is a truism. But I think that many of us would be surprised by how slow many organizations are to notice change when it is sitting in a report right in front of them.
I believe that variance analyses are a sort of “canary in the coal mine” of contact centers—it is an early warning. Variance analyses measure change to the operation, although many execs seem to think it measures forecast “error.” When volumes, or handle times, or attrition, or shrinkage are different than planned, there may be some permanent issue affecting the operation.
Variance analysis is a real simple thing—it is a comparison between what happened and what we thought was going to happen. Small changes can usually be written off as forecast variance, but bigger changes should be considered an early warning that needs to be explored.
If a company does not use variance this way, or is just slow in responding, then we know what happens next: we can get behind the hiring curve and be considerably understaffed (or overstaffed). Both are strategic issues.
Not Planning for Training or Controllable Shrinkage
I do see, periodically, shrinkage that is simply averaged across the year and looks as though it is flat week over week. This is almost always a mistake on the part of the planner, as shrinkage is both a seasonal occurrence (for uncontrollable shrink) and a strategic decision (for controllable shrink). Let’s discuss controllable shrink.
How often do we find our contact centers overstaffed in a given week? When we are overstaffed, it may likely be a strategic planning error—four or eight weeks ago, the long-term planner may have been able to show management that there were weeks where extra training or vacation without pay may have been able to help keep service consistent and the contact center running efficiently.
How often do we find our centers understaffed? Often, we hear stories of management pulling agents off of the phones in order to work on side projects, or planning meetings without regard to the workforce management team’s input. A good planner may have been able to tell management weeks ago that we can expect to be understaffed, and that we should prepare our operations for overtime or other contingencies.
Choosing not to look at the longer term plan is a poor management decision.
Handle Time Forecasts are “Off”
A funny story I sometimes tell is about a large contact center that found their handle times much higher than forecasted, and their service levels well below standard. My friends, the planners, were under the gun to “fix” their forecasts, even though all the math was pointing to normal handle times.
They could not explain the variance, although they tried. It was a stressful time for the planners, and senior management aimed directly at “forecast error,” and the planners, as the culprit.
Well, it turned out that one of the planners overheard an agent discussing a brand new policy, where agents were encouraged to make the customer experience a priority over all operating metrics. Everything started to make a little more sense. Unbeknownst to the planners, a new center manager chose to roll out new KPIs that encouraged agents to speak more to the customers. Handle time targets were no longer a priority.
But nobody looked at the cost of this policy. Developing new policies, without understanding the cost repercussions (or informing the planners that there will be a new policy), is a bad, bad practice.
My all-time favorite planning story was about how a senior director at a financial planning organization asked the planning team to evaluate a hiring freeze. The manager of the planning group used his planning system to show that a service catastrophe would be the result. So the executive asked about half a hiring freeze. The manager showed the boss that they would still have a service failure, just a little bit later.
But he asked the exec, “What are you trying to do?” and the executive told the planning team that he was told he needed to cut costs—two million dollar’s worth.
This is where the planning manager became a hero. He asked for a day to do analyses with his planning system, and returned to the board room the next day to show the executive all the things the contact center could do to reduce costs—without compromising service. Reducing handle times by 5 seconds would save $800K. Moving 2% more calls to self-service would reduce costs by $700K. Reducing shrinkage by 2% would save $500K. This was very cool analyses.
The question, he asked of the executive, is whether you can accomplish these improvements. If you can, you will find two million dollars.
Risk Analyses Can Help!
One of the most powerful types of decision analysis you can perform is to draw out the risks and costs associated with a particular set of decisions. It sounds complex, but it really isn’t. Let’s say we want to evaluate the best course of action after we found our volumes were increasing. What should we do?
Well, we could staff up. To evaluate the cost of staffing up, we would simply add the new volumes to the planning system, and develop a new capacity plan. The costs of this plan are super easy to calculate. But what happens if we don’t staff up? Well, we add the new volumes, but don’t staff up. What does our planning system say about service levels? How bad do they get?
Comparing the two decisions shows the value of making the hiring decisions today. We are showing the trade-off of service versus cost, and by presenting the trade-off to our execs, they can see the operational risk associated with not staffing, as well as the cost associated with staffing higher to accommodate the volumes.
There are so many cool risk analyses we can do, so long as we can evaluate a new capacity plan quickly and accurately, and if our leadership team is willing to let them avoid a bad decision by listening to us.
Ric Kosiba, PhD is a charter member of SWPP and vice president of Interactive Intelligence’s Decisions Group. He can be reached at Ric.Kosiba@InIn.com or (410) 224-9883.