Predictive analytics, the ability to mine, analyze, and model current and historical data to predict future behavior or events, is an old concept made vital by the potential of big data.

As with all analytics, the skills to effectively model the volume and variety of data are in limited supply. Additionally, predictive analytics requires a shift in the call center culture.  The adrenalin highs of reactive management must be replaced with a proactive culture that is focused on the future and on prevention. Fee structures, patient loyalty, and financial health of firms within the healthcare industry will rely upon predictive and prevention focused analytics.

While deploying speech as a self-service channel has been challenging, the application of speech analytics to recorded customer interactions is powerful.

By analyzing recorded calls this technology allows us to correlate agent behaviors and call outcomes to customer preference, generating actionable and compelling insights.

Beyond customer and potential competitive insights, speech analytics plays a significant role in performance management by enabling precision call monitoring, supporting root cause analysis of agent performance failures, and providing specific coaching insights. These benefits do not accrue magically; front-line management will need to be reskilled to leverage speech data for agent performance improvement.

Like big data, speech analytics requires analytic skills and business savvy when mining for customer insights. This is not plug and play technology with standard queries and reporting; it requires a baseline understanding of the customers’ mandatory requirements and service preferences to effectively leverage the tool and structure queries. Organizational biases and prejudices can in fact limit effectiveness when mining for customer or operational insights and thus must be discarded.

McIntosh supports clients as they explore the capabilities and benefits available through speech analytics.  Contact us for more information on how we can accelerate the speech analytics learning curve and achieve rapid payback on your investment in this compelling technology.

For too long, contact center management has labored under the misconception that entry-level supervisors either arrived on the job with the necessary management skills or that those skills were acquired via osmosis.  Neither is true.

Meeting operational objectives requires planning, metric clarity, effective measurement tools, disciplined communications, and accountability for results, both successes and failures.

Performance management in most call center environments is ineffective due to one or more the following five conditions:

• Call centers create a reactive culture focused on managing the crisis of the day.  Management is consumed by daily operational dramas and is unwilling or unable to engage in planning and proactive efforts to prevent daily firefighting.
• Call centers measure everything and thus generate metric clutter.  Performance reporting is excessive, unfocused, and does not present a clear snapshot of the organization or an individual’s performance status.  Coaching is not enabled by excessive data; instead, it is disabled.
• The organization lacks the analytic skills to go beyond identifying a metric deficiency and delve into the root cause of performance failures.  Trend charts which enable proactive interventions are not common at the individual and team levels.  Call monitoring tends to focus on generating statistically valid scorecard data when it should be leveraged as a tool for understanding the underlying causes of high transaction times, poor customer survey scores, or low revenue generation.  Monitoring should generate diagnostic insights.
• Coaching tends to focus on an agent’s ability to meet a target rather than on an educational process leading to performance improvement.  Telling an agent, “Your handle time is too high, get it down,” is coaching to a target.  Explaining that AHT is too high because the agent summarizes excessively is coaching to a skill that leads to performance improvement.
• Supervisors and managers must abandon the concept of a performance plan as the first step in a disciplinary process and treat performance plans as vehicles to instruct and close skill gaps.

At the close of each day, most call centers know to the thousandth of a second how long customers waited in queue before their calls were answered and the length of the longest and shortest wait times.  But, they cannot tell you how many customers were satisfied, disillusioned or pleasantly surprised with their calls that day.  They can tell you on average how long an agent talked, but they cannot tell you with statistical certainty which agents performed well by servicing customers effectively while complying with the company’s policies and procedures.

The ability to measure and report performance is no substitute for understanding performance drivers and coaching to the root cause of performance failures.

Effective Performance Management

The critical success factors that must be met to design and deploy an effective performance management process include:

• Redefining roles and responsibilities so highly effective performance managers are accountable for their teams and the agents on their teams.  In order to do this, we need to begin with a time and motion study that clearly identifies the workload that must be distributed to enable supervisors to focus on performance.  A top priority is the deflection of low value administrative tasks to other parts of the organization, thereby freeing up sufficient supervisory time for performance management.
• Training front-line supervisors to create daily plans, communicate effectively, analyze performance gaps for root causes, and coach effectively.  This training is not a simple classroom exercise but rather is reality based.  It includes traditional instruction yet relies on mentoring and modeling for more than 80% of the learning process.  Training should range from ten to twelve weeks to ensure that the learned skills become effective habits.
• Automation enables performance management through the generation of relevant, daily performance scorecards that provide direction for root cause analysis deep dives. Firms without scorecard reporting should look at interim solutions while the “perfect solution” is in development.  Most importantly, firms must re-evaluate performance metrics, focusing on shared business and customer experience goals to create a balanced scorecard that is uncluttered, enables accountability, and provides direction for root cause analysis. We must redefine the performance management process to rely on results, not the excessive documentation of activity that may or may not generate results.
• Senior leadership must support the new performance management processes and tools and must be highly visible in communicating their support.  Performance management should not be regarded as just another fad that will pass but rather as the new approach for how work is accomplished within the organization.  Following the process must be mandatory if the new approach is to be sustainable.

McIntosh has a proven curriculum and approach and supports clients in the design and delivery of sustainable performance management processes. For a summary of our approach, see our Performance Management Curriculum.