Measuring Agent Productivity In A Multi-Channel World
To manage the contact center well, supervisors must monitor a range of KPIs to measure agent performance and productivity. But use of multimedia (email, chat, SMS, etc) throws up many challenges to accepted methods.
Take for instance the question “Are my agents wasting any time?” In the voice-only world, it’s pretty easy; just measure idle time (or subtract talk and wrap time from available time).
However, the timing of packet media activity (i.e. specifically email) is fundamentally different to that of streaming media (i.e. voice, video). This makes it much more difficult to measure lost agent productivity.
Why? What’s the difference?
With voice/ video, you can pretty much guarantee that from the start of the conversation to the end, the agent is constantly engaged.
Imagine two neighbours, Mr. A (the agent) and Mr. C (the consumer), in houses on opposite sides of the street.
With streaming media, both Mr. A and Mr. C open their front doors and a face-to-face conversation ensues. We can see that from the moment Mr. A opens his door, he is busy in conversation until he shuts it again.
In this case, if you keep an eye on the length of talk/ wrap/ idle times, you can be assured that your agent is productive.
But with packet media, a time delay factor is introduced:
Mr. B opens his door and hurls a message through Mr. A’s window. After some time, Mr. A reads the message, considers a response, does some research, consults colleagues, makes some tea, writes a response, edits the response, opens his door and hurls his message through the window of Mr. C.
Unlike a voice conversation, between each of these stages, time may pass – could be 1 sec, could be 3 mins – perhaps unwittingly, perhaps on purpose.
During those moments, the agent could be doing anything, from engaging in other workflows – emails, chat sessions, text conversations – to watching YouTube. We just don’t know because we are not measuring it.
Under these circumstances, unproductive agent time becomes harder to identify, let alone measure.
What can be done?
Here are 3 possibilities:
- Get a 360 degree view of agent
Because multimedia work is fragmented and non-linear, supervisors need a unified view of all agent activity across all media channels. Work should be viewable chronologically, by media channel, by customer, and actually, by pretty much anything you choose. This does entail agents signing onto and off from activities, a practice that is not commonplace right now. - Develop workflow metrics
Detailed metrics such as time to reply should be augmented by longer term workflow metrics, such as completion rate and number of incompletes. As time passes and outstanding workflows remain unresolved, reporting systems should use alerts to bring them to the fore. It is up to supervisors to press for closure. - Change the focus
Augment the detailed metrics with ‘soft’ productivity measures such as customer satisfaction. Consider also that among the top factors in achieving this are happy agents – those that enjoy their work, have a variety of tasks and challenges, and are encouraged to do their best. This becomes more important as basic interactions are increasingly routed toward self-service and automation and agents spend more of their time in higher value, more complex transactions.
As the multimedia contact center matures, management and measurement must move with the times. Consideration must be given to those things less easily measured, e.g. quality of response, customer satisfaction, as well as traditional, easily-measurable metrics e.g. speed of response, completion rate. Contact centers must – as always – strive to balance of quality and quantity.
