White paper for Enkata workforce analytics software company

Call Reasoning: The First Step Towards Satisfied Customers and Lower Costs

 

Introduction

What are the top 10 reasons your customers call?

Don’t know? You’re not alone. Most companies don’t. Unfortunately, not knowing is costing you more that you think. When you do understand why your customers are picking up their phones and dialing your number, you can take corrective action, addressing the issues that cause high call volume. You can:

• Optimize your self-service channels, training and coaching.
• Fix the problems in upstream processes that end up in your call center.
• Reduce repeat calls.

In short, you can save millions of dollars each year in call center and customer defection costs, and ultimately improve customer satisfaction.

Call reasons are a powerful tool. But, identifying them hasn’t always been within reach for most companies. The process has historically been difficult and expensive, and the reasons produced have been unreliable.

A common approach saddles agents with the responsibility to tag their own calls, a process that yields inaccurate, incomplete data and costs an extra two to four seconds per call. The category choices may be too limited to express the range of issues covered in an agent-handled call, or may exclude choices made by the customer prior to talking to the agent. Agents may choose the first available option, or the broadest, as they rush to wrap up a call and move on to the next.

Other methods, including post-contact surveys, quality assurance logging, and speech analytics, collect information from a random sample of calls. Sampling doesn’t provide enough data from enough calls to be reliable, or the granularity required to tie specific issues to individual agents. Speech analytics applications don’t identify call reasons at all, just keywords.

Additionally, most of these methods only allow one reason per call, when in reality there could be multiple reasons. To further complicate matters, there has been no single, consistent method for assessing contact reasons across channels, including web interactive voice response (IVR) and agent-handled calls. For example, one customer may attempt to resolve the same issue using multiple channels, but because each channel calls the reason by a slightly different name, it will end up looking like three different reasons.

The good news is that all the unsuccessful methods have created a tremendous opportunity for improvement.

Enkata recently introduced analytics technology that can determine call reasons by categorizing the patterns of customer, agent and system activities. The result is an affordable, accurate categorization of 100 percent of your calls. (Not to mention the two to four seconds of average handle time [AHT] you’ll save on every call that agents don’t have to tag.)

Enkata can deliver more accurate call reasons, which give you the power to proactively address the issues that cause your customers to call in the first place. Knowing your call reasons can potentially help you improve agent training and coaching, revise cumbersome policies and procedures, and save millions in operational costs. All these improvements ultimately funnel into better quality service and higher customer satisfaction.

Enkata’s Automated Call Reasoning solution can give you the reasons you need to truly optimize agent performance and increase the productivity and effectiveness of your entire operation.

By knowing their call reasons, Enkata customers have:

• Improved new hire training effectiveness by 50 percent
• Increased coaching effectiveness by 50 percent
• Improved self-service success by 15-30 percent
• Reduced AHT (related to tagging) by 2-4 seconds
• Reduced unwanted call volume by 2-5 percent

Why Automated, Algorithm-Based Call Categorization Is So Accurate

Enkata’s unique Automated Call Reasoning solution categorizes all of your calls automatically—and very accurately, because Enkata assigns reasons based on the actions your agents take while they’re handling each call. We developed this technology to assign consistent, reliable call reasons without burdening agents or other internal groups with the task. Once you’ve installed Automated Call Reasoning, you can begin to understand why your customers are calling and which call reasons cause your agents to struggle in a matter of months. Here’s how.

 

During every call, an agent accesses multiple systems to look up information and perform transactions on the caller’s behalf. The agent’s activity within each system contains clues that Enkata uses to determine why the customer called.

Enkata’s Automated Call Reasoning solution uses an algorithm to combine data on what workflow the agent followed, which back office transactions occurred as a result, what information was entered in customer relationship management (CRM) systems, etc., and produces a standard, hierarchical set of reasons for every single customer contact. Enkata’s method truly provides the most accurate call reasons, because we derive and verify customer intent by using all of these reliable sources. Read more below about what we gather from each system.

Agent Desktop Systems
The primary source Enkata uses is the logs, or the records, from agent desktop systems. These are very detailed and reliable, as they indicate the precise navigational flow of the call by tracking the screens viewed and specific activities performed by the agent. (See illustration above.) A call typically includes two or more “screen change” events that we track. There’s a standard verification screen that starts the process, an end-of-call cross-sell attempt or wrap-up screen, and then any number of screens in between these two. It’s these in between screens that are the most valuable in deriving the call reason or reasons.

Back Office Systems
When the agent front-end data is insufficient or not available, Enkata’s Behavior Analytics can use a back office system to determine caller’s intent. We can often reconstruct a caller’s intent by reviewing related events that took place around the time of the call. In the absence of richer data, Enkata assigns a reason based on the transaction type of the event (fee reversal, address change, etc.) that follows the call.

CRM
Alternatively or in addition, we can use a CRM system that keeps a track record of calls. It’s not an ideal scenario, because CRM systems don’t provide as much detail, and also because agents typically fill in the customer intent manually. Additionally, some CRM applications include a system-generated call reason in addition to the agent-populated one. When available, this data helps to determine how frequently the two reasons differ. By reviewing the call reasons from both sources, you can see which is more reliable, or find a way to use both sources to arrive at true customer intent.

 

Additional Channels
Enkata’s Automated Call Reasoning module can also use data from other contact channels to produce consistent reasons between calls and self-service options. Additionally, information from other channels can help determine the volume of calls that could be handled by self-service channels. Enkata can use data from customers’ unsuccessful attempts in other channels to resolve their inquiries to determine their true end-to-end experience.

Enkata’s Algorithm-Based Methodology to Determine Call Reasons

Enkata provides a structured yet flexible methodology to generate detailed and accurate call reasons that are tailored to your business processes. Even after deployment, Enkata continues to monitor and update call reasons as operations change.

Step 1: Determine inputs and capture data
First, Enkata works with the customer to determine what data is available from which systems. Enkata uses these to generate a standard set of call reasons based on agent activity. From there, we look one level of detail down to understand the data sources and specific agent activities that will reveal the reason behind each customer contact.

Step 2: Eliminate input “noise”
This is a very important step, and one of the main benefits of using Enkata’s methodology. Our algorithm is uniquely able to produce trusted reasons, instead of just listing transactions, in part because we can pro-actively tell it to disregard certain events that you don’t want to influence the derivation of call reasons. For example, you might choose to ignore an agent activity that is performed for every customer contact, such as account verification or checking the customer’s balance. We identify these “ignore rules” for each data source.

Step 3: Generate raw events table
Next, we take the data derived from step one and remove the transaction specified in step two. The result is a table of 25,000 to 30,000 uncategorized, or “raw,” transaction event combinations. It’s the launchpad for determining how we’ll map different transaction combinations to contact reasons.

Enkata uses this same process for every data source (phone, web, etc.), which is how we ensure you get consistent call reasons across different contact channels.

Step 4: Trim category list and map to tailored hierarchy
Through an iterative process, we eventually collapse the list of event occurrences from 25-30,000 to about 125-150. We start with Enkata industry templates for call reason hierarchies, map those reasons to the raw events, and then test by listening to calls and re-mapping as necessary. We repeat this process until we get to right level of detail with the call reasons. Enkata and the customer work together to define this tailored set of reasons and the number of hierarchical levels desired for analyzing contact data.

Step 5: Assess reasons for every call
Upon deployment, your customer contact data gets loaded into the Enkata datamart, and the Automated Call Reasoning algorithm assigns every call one or more call reasons (up to 3) based on the activity associated with that customer interaction.

Deliver Higher-Quality Customer Service

Illuminate the issues that plague your service center. Automated Call Reasoning tells you why your customers are contacting you, empowering you to proactively address the issues behind high call volume. Reducing the total number of calls, as well as improving your response, soon adds up to cost savings and increased agent and customer satisfaction. Enkata’s Automated Call Reasoning gives you:

 

• Reliable call reasons
  Accurate and actionable call reasons across channels give you what you need to analyze and improve your contact center’s quality of service.

• Money and time savings in call handling
  Because Automated Call Reasoning automatically tags a call based on agent activity, agents don’t need to waste the 2-4 seconds per call to manually assign reasons.

• More effective, targeted coaching
  Armed with call reasons, supervisors can make the most of their limited coaching time. Call reason level details on key metrics gives supervisors the details they need to focus their training and coaching activities on the specific call reasons that cause agents to struggle. Informed supervisors save time and make a bigger impact.

• Increased agent performance via actionable metrics
  When agents know which calls give them trouble, they’re more likely to self-correct, seeking the help or information they need to be more efficient.

• Money and time savings in call auditing
  Many contact centers spend a significant amount of time and money to audit small samples of calls and check the accuracy of agent-tagged call reasons. Enkata’s Automated Call Reasoning solution obliterates the need and costs for these antiquated procedures. Automated Call Reasoning provides more reliable reasons for every call at a lower cost.

• Improved customer satisfaction
  Ultimately, you want to know why your customers are calling so that you can solve their problems before they call — by finding and fixing the defects, processes or policies that cause high call volume. And if you can’t do that, you want to help them as efficiently as possible once they do call. Enkata’s Automated Call Reasoning gives you the information you need to identify where to concentrate your time and attention. When you know why customers call, you can specially train different groups of agents to handle specific kinds of calls. You can individually coach agents who struggle with certain reasons to perform better. In the end, it all comes down to providing high-quality, efficient service, which will always translate into higher customer satisfaction.

Conclusion: Call Reasons Pave the Way to Better Service

Call reasons are a powerful tool. Most companies don’t know the reasons why their customers call, because identifying call reasons has historically been difficult and expensive, yielding unreliable results. But that’s no longer true thanks to new technology. Enkata’s Automated Call Reasoning uses an algorithm-based method to categorize 100 percent of your calls—accurately and automatically. That means authentic reasons with no manual tagging required, without reliance on random samples or keywords.

Enkata accesses data from agent desktop systems, back office applications, CRM products and other channels to map agent activity and service events to call reasons. We then apply our proven methodology to sift through the data, prioritize it, and assign up to three call reasons per customer contact.

Get out of the cycle of handling avoidable calls and wasting money on call center costs and lost customers. Armed with accurate call reasons, you’ll have insights into how to improve your upstream processes, self-service, training and coaching so you can take action. The knowledge you gain from Enkata’s Automated Call Reasoning can help you transform your organization through more productive supervisors and agents. Knowing your call reasons is the first step towards truly optimized contact center performance, cost containment and improved customer satisfaction.

 

About Enkata
Enkata helps contact centers maximize the value of every customer interaction, deliver better service, control costs, and generate revenue. For customer operations transitioning into strategic profit centers, our performance management software leads the market. Enkata’s unique on-demand delivery model allows for faster implementation and results. Our analytics-driven applications combine best practice metrics, personalized dashboards, “next action” direction and integrated workflow that improve the balanced achievement of sales, customer experience and cost objectives every minute, every day.