A leading health provider with over 6,000 CSRs found that the underlying cause for wide variation in AHT among agents (a 45 second difference for certain call types) was the use of quick keys. Quick keys. The training of ‘slow agents’ on quick key usage resulted in a $3 million annualized savings.

Ask any supervisor for an agent’s statistics on average call length, longest call, number of calls handled per day, wrap up time, etc., and they have the data at their finger tips. Ask any supervisor questions about agent typing speeds, data entry error rates, time spent in various applications, time required to complete a transaction, or complete a step in a transaction, and you will likely get a blank stare.

Ironically, an untapped, rich source of behavior data has been staring CSRs in the face all this time – their desktops. But there is near zero visibility into CSR desktop activity, primarily due to lack of systems or tools to effectively capture data on desktop activity and present it to supervisors with meaningful, actionable coaching indicators. An emerging class of desktop analytics software applications promises to automate these analyses across multiple CSR behavior types.

Seven desktop behaviors you should track and coach out of agents
Monitoring and coaching against the following desktop behaviors provide quick returns on coaching investments and support the identification of, and training on, best practices.

Skill Proficiencies:
1. Typing factors: typing speed, error and inaccuracy rates, backspacing and retyping.
2. Data entry practices: quick key usage, overuse or underuse of notes, copy and pasting.

Practices:
3. Application navigation: efficiencies associated with scrolling, resizing, navigating application to application and within an application, use of navigation jump keys, ability to multitask across multiple applications following efficient sequences.
4. Application misuse: frequency of mis-opened applications and identification of those applications, use of rogue or alternative applications that improve or inhibit productivity, excessive time spent in specific applications (e.g., knowledge databases), excessive time spent searching for required forms and/or information.
5. Application abuse: time in non-compliant applications such as social networking sites or restricted applications.

Compliance:
6. Process compliance: monitoring for compliance on data entry, transaction steps, transaction workflows and number of manual overrides.
7. Security compliance: logon and logoff violations; fraud violations; access to restricted sites, applications, and windows; length of time sensitive windows are left open or in active window state.

What you can do about it
Coaching programs tied to emerging analytics-based software tools will drive a new set of performance improvements at the agent level but also deeper operational performance improvements.

The first factor in achieving performance improvements for these desktop behaviors is the ability to monitor CSR activity. However, unlike the sampling approach for quality monitoring of phone calls, desktop analytics will monitor and capture all customer interaction data from the desktop and analyze 100% of desktop sessions (customer calls).

Factor two is the ability to measure and quantify the data for agents relative to their peers. With this data, contact center managers can leverage the best practices of top performers and identify agents in the lower performance range that require coaching. A critical component in any solution is the ability to categorize “work types” – to create apples to apples comparisons of agent behaviors.

The third factor is the ability to identify specific issues and skills on which an agent should be coached. This supports a clear and personalized coaching plan that is mapped across the full spectrum of desktop behaviors.

The fourth and final factor is the ability to measure and quantify performance improvements over time and more importantly to tie those improvements to training and coaching plans as justification for coaching investments.