Spikes in Call Volume – How to Stay Balanced with Real-Time Workforce Management
“We are currently experiencing unusually high demand. Please try again later”.
Not exactly an excellent customer experience.
Organisations looking to retain customers will try to play that message as little as possible. That’s why you put so much effort into Workforce Management (WFM), scheduling just the right number of agents for the expected workload; not so many that you’re paying staff to twiddle their thumbs; not so few that they get burned out and leave.
But then… the unforeseen. Maybe it’s a spike in call volume on one of the queues. Customer wait times sky-rocket and supervisors scramble to figure out where they can pull talent from, without wrecking other queues.
Or maybe there’s an outbreak of Covid and suddenly you are understaffed.
What you are aiming for is balance, peace and harmony between all workloads so that no single service level dips (or plummets!), outreach continues without compliance trouble, and all agents are working at their comfortable peak.
This problem of real-time management is like flying a drone. You take into account wind speed and direction, trees and other obstacles (and the occasional bird). One drone: easy!
But how about flying two at once? Not so easy. Maybe you get a buddy to help, but to avoid collision you have to coordinate your actions. The more drones you need to fly, the more buddies you need, and the harder it is to keep the whole operation in the air.
So it is with queues and supervisors; the more workloads there are, the more supervisors are required to manage them, and the harder it is to maintain balance without collision.
So, if balance is the goal, how can it be achieved?
Let’s look at three possibilities for real-time (intraday) workforce management:
The record-breaking drone light show at the 2018 PyeongChang Winter Olympics, featuring more than 1,218 drones. (Photo: Intel)
1. Manual real-time workforce management
In this scenario, supervisors constantly monitor queue KPIs, and take action to reassign agents if any service level is under threat.
Downsides with this approach:
- Negative effects on other workloads
What will be the negative effects? And how severe will they be? The answers rely on finger-in-the-air, intuition and guesswork, leaving it open to human error. And mistakes can be costly.
- Performance decreases with scale
This approach may work for a few queues, but quickly gets difficult as you scale up. More queues means more supervisors, who are more likely to trip over each other in the scramble for balance.
- Supervisor cost increases with scale
Supervisors are busy with unnecessary (but costly) work. And the effect compounds: the more queues there are, the more supervisors are required to manage them.
- Competing priorities
Supervisors may simply not allow ‘their’ agent resources to be poached, as it would have a negative impact on their own teams/ workloads. Supervisor incentives must be carefully balanced with company objectives to avoid competition.
The manual approach may be OK to manage a small number of queues. But the more there are, the more priorities conflict, the harder it is to maintain balance, and the more costly it is in supervisor time.
2. Rules-based real-time workforce management
In this scenario, rules are set to trigger particular actions when thresholds are met, perhaps assigning different skills to certain agents, perhaps overflowing to a BPO service.
Downsides with this approach:
- Advance setup
Rules need to be set up in advance. Actually, what you would like to happen when trouble strikes is harder to predict in advance, than in real-time response.
- Real-time monitoring
As with manual, supervisors are still tied up with monitoring to ensure that things behave as expected and don’t cause more problems. And they still have to step in if things go awry.
- On/off switch
A simple trigger threshold offers no granularity, and therefore no smooth operation. Queues can rapidly fluctuate above and below threshold, especially as queues are played off against each other, leading to unpredictable, sometimes bizarre behaviour.
As with manual, the rules-based approach may be OK to manage a small number of queues. But the more there are, the more priorities conflict, the harder it is to maintain balance, and the more costly it is.
3. Automatic real-time workforce management
In this scenario, agents are reassigned automatically by the ACD from spare capacity or less urgent work to help with any service levels under threat, and calls/media sessions are assigned automatically according to agent capacity, skills etc. When making decisions, all factors are taken into account – multi-channel capabilities, skills, SLAs, wait times, potential abandoned outbound calls.
Upsides with this approach:
- Better balance
Performance is balanced across all workloads, allowing no catastrophes on a single queue
- Better scaling
Performance increases at scale. As the entire system is connected, support can be mustered to lend a hand wherever and whenever necessary. The more queues, workloads and agents are available, the more any increased load is shared between them. This makes best possible use of all available resources.
- Better utilisation of supervisors
Supervisors are freed from managing agent movements/ skills and can concentrate on getting the best from their team.
Automated intraday workforce management maintains the best possible balance across the enterprise, even at scale.
It’s all about maintaining balance; being zen, letting the system take the strain and make better decisions than humans can.
To fly multiple drones, the solution is the same: automation, which can achieve and maintain far more complex synchronisation than is possible with humans. For example:
Imagine all inbound queues, outreach, customer engagements, workloads and compliance needs being balanced beautifully for you. No sweat, no strain, no stress…