How Workforce Management Decisions Can Ruin Everything

Original Publication: Customer Management Insight - October 2008
View Online

If you surrender your contact center to the tools designed to help you run it, expect mayhem.

I will now share with you one of the two fastest ways to run your call center straight into the ground. It’s called “following your gut.” Oh, you might get lucky once in a while with your gut, but a steady diet of following it will — absolutely, 100 percent guaranteed — be the business equivalent of sticking your hand into a wasp’s nest.

An example of what I mean: I once went with my gut and convinced myself that my cat was fine and dandy. “Look at how lively she is,” my gut told me. “If she had a bladder infection. You’d know.”

A week later, the vet, practically laughing as he read out the test results over the phone, knew, too. He knew a lot: what kinds of crystals and infections form in cat bladders, how much to charge for the exam, for the medicine, for the special cat food. The man was a walking encyclopedia of pricing information. And he knew it.

Workforce management is particularly prone to the same sort of “gut” instinct mentality and the gently rippling pool of lunacy that the gut will drag you into — kicking and screaming. The way around the humiliation and despair that will ensue in this disaster of your own making? Like almost everything else at call centers, software offers a rope long enough to pull you back to the sandy shore of good call center management.

Although the sophistication, feature sets, and costs differ between vendors, almost all WFM packages do the same things:

1. Collect data from automatic call directors (ACDs), multimedia skills-based routing (SBR) packages, and other relevant systems.

2. Forecast contact demand. In virtually all cases, a well-tuned WFM forecasting module will produce more accurate results than a call center manager’s or scheduler’s intuition.

An interesting entry that uses both formula-based and simulation-based forecasting is AgentTime Scheduler from Portage Communications in North Bend, Wash. Portage Communications sells WFM software primarily to small- and medium-sized call centers.

According to Paul Leamon of WFM software developer IEX in Richardson, Texas, one trend in WFM software is gaining “capabilities to handle the ‘three multi’s’ —  multisite, multiskill, and multimedia.” You’ll need a WFM package that addresses your company’s needs in these three areas.

Note also that most call center WFM packages have limited capabilities for the scheduling of outbound contacts. Concerto Software in Westford, Mass., is the market leader for outbound call center WFM.

3. Create work schedules that balance anticipated workload against agent availability, shift flexibility, costs and service level requirements. These schedules can include vacations, lunch time, breaks, meetings, and coaching or training sessions.

4. Assign agents optimally. This requires human judgment. A purely top-down process is almost certainly going to be in conflict with your long-term goal of retaining good agents. One trend is to give points to agents based on seniority, performance levels and skill sets, then have the WFM system assign agents by bid order preference to the available schedules, subject to constraints that the center’s WFM analyst specifies. But don’t be too doctrinaire.

5. Provide information that lets supervisors, managers and analysts engage in day-to-day management.

6. Monitor agent adherence to schedules in real-time and intervals. Be realistic though. The likelihood of getting perfection for extensive periods in your schedules is pretty low. Again, don’t be too doctrinaire.

Rajiv Venkat of Witness Systems in Atlanta, Ga., suggests that software that lets supervisors see adherence patterns over time, as well as in real time, can help avoid “Big Brother” behavior by supervisors. He says this “changes supervision from an authoritarian approach into coaching.”

7. Facilitate communication about scheduling. If you’re looking at new software, you should examine whether it lets you specify permissions for groups and individuals (e.g., allowing each supervisor to approve schedule changes only for the agents that he or she supervises).

8. Estimate the financial or service-level impact of operational changes using “what if” scenarios.

No! This Is Major Surgery!

Years back, I happened across an infomercial for an elective surgical procedure. The host asked the spokesperson who was marketing the procedure something like, “So, is this performed as an outpatient procedure?”

The spokesperson’s response (I paraphrase) was immediate and surprisingly forceful for the usual mirth-filled, light-heartedness of informercials: “No! This is major surgery!”

In the same way, introducing WFM software is not a quick little out-patient procedure done in your doctor’s office under a local anesthetic in 20 minutes. This is major surgery. The amount of change for agents, IT, and the management team is a challenge — and by “challenge,” I mean “potentially disastrous situation” — for even the best-prepared team. A failed implementation will haunt the organization with a directionless workforce planning team, unhappy agents, frustrated managers and a system that no one uses.

Many organizations tend to go into the process unprepared. Why? For the same reason there are about a thousand abdominal muscle machines out there and maybe 17 people with (let me find the exact phrases) “a tight, sexy core,” “rock-hard abs,” or “an 8-pack.” People believe the marketing surrounding the purchase of WFM software or an Abdominizatrontor Mark XIV — that just the action of buying it will make everything all better.

And because you bought a system just on the strength of the marketing, each setback is compounded by an increasing lack of trust and disappointment. Soon, the entire implementation is in jeopardy, and you’re under your desk crying quietly so none of the others will hear you. Any (employed) executive finds that a scary proposition. To avoid it, ask yourself some questions before you go shopping.

Do We Have A Plan?

Generally, when you watch a demo of a WFM solution, the system forecasts, then it builds schedules and plans for off-phone time, and finally, you end up with great service levels (to go along with your washboard abs). Julia Child would have been the perfect WFM solution sales agent. Take this. Do that. Stir this. Butter that. Voila! She got a soufflé; I got an oven fire. And just like Julia Child, the wonderfully misleading WFM visual makes everything look so easy! I’m convinced, absolutely certain, I can put that caramelized crust on the crème brulee with a blowtorch. I’m convinced, absolutely certain, I can drive myself to the hospital to have those second-degree burns on my hand tended to before I pass out from the pain.

WFM software facilitates the process, but it cannot make it happen. If you forget that, you’re doomed. Not “in for a challenging opportunity.” Doomed. Like the Titanic. Plain, old, standard, regular doomed.  

Do We Have A Good Plan?

Julia Child could put together a meal that would have made you weep. I can do the same thing, but you won’t be weeping for quite the same reason. For WFM software to help with your center's staffing, you have to have a good plan.

The primary component of a successful implementation requires the habit of having a planning culture in which all off-phone time is scheduled, not two minutes beforehand, either. You must already know what you’re doing; you have to already be forecasting, scheduling and planning off-phone time. A center with this culture has a defined service level goal, sufficient budget to staff it and an organization-wide focus on service level as a key metric. The call workload is forecasted accurately and the staffing assumptions validated. Finally, all off-phone work is planned and integrated into service level forecasts.

So where’s the benefit in the WFM software usage? The automation of time-consuming (i.e., boring, repetitive, repetitive, repetitive) processes, such as scheduling and break assignments. That frees up time and energy for more thoughtful, in-depth analysis of schedules and forecasts. Julia Child didn’t chop onions. She gestured at a bowl filled with onions chopped for her by off-camera minions. The WFM software is your off-camera minion. Use it for the scut work and grab your blowtorch and soufflé pans and get going.

Who Do I Need To Run This Thing?

Workforce planning is a detailed and analytical process that is dramatically different than most other call center processes. A call center will need at least one dedicated person to run the WFM system, handle manual tasks like data entry and analysis, and manage the planning process.

Historical call and handle-time data must be cleaned for system outages and other anomalies. Forecasts must be adjusted for non-historical events or call drivers. Scheduling change requests and off-phone time must be entered. Rosters must be maintained so that the correct schedules are generated. Schedule sets must be created. Breaks, lunches and meetings must be analyzed to maximize service level.

This is a difficult reality for centers that are just large enough to consider buying WFM software. The added expense of having an on-staff someone run the software often is not in the budget. In many cases, supervisors or part-time resources are handed the task. The result? Something — or someone – breaks down, and you get an under-utilized system that cannot produce the desired results.
 
Larger centers transitioning from spreadsheets to a commercial WFM system often discover that they see minimal benefits from the implementation after the first schedule run. Initial service level gains often level off with little improvement thereafter. Generally, this is due to a training issue with the WFM team, such as employees who are used to data-entry work and basic analysis now have the time, but not the skills nor knowledge, to perform more advanced analysis. In most organizations, a position is eliminated, or the lack of skills is ignored. Training in advanced statistics, MS Excel or Access, forecasting methods, and the visual presentation of data has proven valuable for improving the analytical output of WFM teams. These skills allow them to take data and turn it into the analysis managers want. You have to consider retraining schedulers on how to analyze and present data. That will lead to better reports and smarter staffing decisions.

Slices, Doesn’t Dices

Although it’s a possibility, the reality is that no single call center does everything right. (Except yours. Yes, yes. Please don’t send me death threats.)

I put the question to Ron Hildebrandt, founder of Enkata, a provider of call center performance management solutions: “Is there a glaring, obvious functionality that a lot of call centers are NOT using even though it’s available?”

“Although first-call resolution (FCR) is considered the most critical metric to manage a contact center, most contact centers do not coach to improve it,” says Hildebrandt. “Because FCR is mainly measured by surveys today, supervisors do not have a metric that is trusted, or enough examples of calls that were mishandled, for coaching on FCR to be effective. The result: Our Number One metric in the call center gets little attention during regular coaching sessions. It is time to automate the FCR metric calculation to enable supervisors to coach to impact it.”

The Second Way To Ruin Everything

It would be gauche to leave you all hanging, so here’s the second of the two fastest ways to sharpen your call center’s heels and drive it straight into the ground. Jane Goodagent — the third-best agent in your center of 200 people — comes up to you. She needs Thursday off. You put down your coffee mug (the one that says “World’s Greatest Boss”), and you lecture Jane for 10 minutes about the “clearly established” procedures that require six weeks’ notification for time-off requests. To make sure the saucy little upstart knows her place, you take out the record of how many times Jane, compared to all the other agents, has used the bathroom in the last fiscal quarter. Wave that in Jane’s face. Remember not to scream. Once you scream, you’ve lost the argument.

End result? You’ve reduced Jane to a crying, bitter wreck. Well done! Your call center is now that much closer to failure thanks to your hamstringing of one of your top performers. You’ve earned a second cup of coffee, boss.

All those machines and bits of software are not Oz, the Great and Powerful, Granter of Wishes to the Wretched of the Earth. You can’t tell software that one of your so-so agents, regardless of what the in-place system rules say, deserves to have the day off. That requires empathy. Empathy is not the same thing as going with your gut. The “gut” jumps to a conclusion in a fraction of a second and leaves you hoping everything will somehow come out satisfactorily in the end. Empathy is the practical voice that whispers, “How would you feel if she were treating you like this? You make these people hate you, and your center’s KPIs are going to drop like a dead bird out of the sky.”

The machines have made a whole lot of things a whole lot easier. But never think they can do it all. They can’t. They never will. That’s what you’re there for.

What’s With All These Bells And Whistles?

One of the latest bits of whistlery to come out of the car industry is the air-conditioned glove compartment. For when you drive out to a sushi place in the middle of the desert (I’d love to see that business plan). Just slip that tray of raw fish into the air-conditioned glove compartment and drive away, off to new adventures.

“Does this car have an add-on that will keep my raw fish purchases cold?” is not something I suspect many people ask at the dealership.

Keep an eye out for the similar “add-ons” in WFM software. Are the what-if scenarios going to be useful if they take all night to run? Do functions freeze when you multitask? There’s no undo? A whole lot of consultants’ children went to college on these issues.

The schedule adherence module provides a good example of how to determine if an add-on is right for your center. Because schedule adherence is such a critical measurement in a call center, it seems logical that every center should have it. However, it really depends on three considerations that apply to any WFM software add-on: frequency (how often you really need to look at schedule adherence), complexity (the more complex, the more likely that you will need an adherence module), and value (is it worth it, in real dollars, to know your exact adherence). Consider all three factors equally to determine if the cost of the add-on negates the manual work it would otherwise take to perform the task.

With any add-on, but especially with schedule adherence, it's critical to calibrate the results with the payroll system. A schedule adherence rate of 92.5% is only good if you don't have missing payroll hours. Ultimately, WFM software is for matching workload to payroll.

Regardless of which add-ons you choose, or how smoothly your implementation and training goes, something will go wrong. WFM software is set up to integrate with a standard installation of a phone system and a standard workforce planning process. Most call centers try to bend their phone systems and WFM processes to meet their business needs. Go into the coffee store and order a latte, and you get a latte. Go in and order a double mocha heated to 183 degrees served in a blue Styrofoam cup and sprinkled with Norwegian cinnamon by a left-handed barista wearing an eyepatch. You end up getting escorted from the building. (I learned that the hard way.)

WFM software is not designed to handle every possible contingency. Speak to the vendor about the things you will need the software to do before you even start thinking about filling out the purchase requisition.

Unfortunately, some cases defy a WFM system's ability to customize. This typically occurs when business processes become convoluted. Overly complex routing or shift design can cause what-if scenarios and multiskill-routing scheduling to slow to a crawl. If your WFM system is not performing as expected, consider how your rules may be contributing to the problem. If the WFM vendor cannot accommodate your rules, the WFM system may not be doing you as much good as you think, or you may be doing more harm than you thought with a bizarre set of conditions that are impossible to automate efficiently.