PATENT TITLE / NAME

Apparatus and method for multivariate prediction of contact center metrics using machine learning

WHAT DOES THIS MEAN FOR OUR CLIENTS

Swampfox knows that customer experience lives and dies based on setting customer expectations and meeting them. Historically, understanding the impact of caller volumes, agent staffing, handling of different types of calls, and responding to scheduled and unscheduled events has meant that setting expectations has been extremely challenging to clients.

Swampfox has now patented the ability to predict many of these core KPIs including the “estimated wait time” for an agent to become available for service using its own approach to machine learning.

Swampfox now makes use of this super accurate prediction capability to set expectations about how long “customers will be on hold” for an Agent and then offer Callbacks via its industry leading Intelligent Callback Solution, First In Line™.

This means that our customers can be absolutely confident that when First in Line tells the customer it will be “20 minutes” before an agent will be available, that First In Line will reach that customer when they say they will. This allows customers to build confidence in the solution and plan the time they would have used waiting on hold.

Therefore, they will be able to make the best decision possible whether they would like to schedule a callback for a later date and time, have an agent call them back as soon as one is available or wait on hold for the next available agent.

With an accurate estimated wait time (EWT), customer satisfaction increases, callback trust and enrollment trends upward and customers can drastically improve call management while reducing staffing and infrastructure costs enabling your organization to thrive.

It also means that Swampfox ICX™ can use this information to accurately choose the best means of resolving a customer need by delivering the customer to a skilled in-house agent, an outsourced agent, to deliver them to an Intelligent Virtual Assistant, to move the customer from a Voice Channel to a Chat Channel, or to offer the customer the ability to enroll for an Intelligent Callback.

ASSOCIATED SWAMPFOX PRODUCT(S)

• First In Line: Predicting, Offering, and Honoring Callback Wait Times
• Swampfox ICX: Predicting and Selecting from Voice Agent, Outsourced Agents, Chat Agents, Intelligent Virtual Assistants, Chatbots, or Callbacks using accurate predicted Statistics including Callback Wait Times, Predicted Agent Occupancies, Predicted Service Levels, Predicted Containment or FCR rates.

ABSTRACT

Our latest enhancement to our First in Line Callback Solution has been specifically targeted at calculating the Estimated Wait Time (EWT) or the time that it will take for an agent to become available to assist a caller. It is imperative to get this calculation correct for a multitude of reasons: to set your customer’s expectations, to accurately forecast and staff appropriately, and to inform your customer accurately so that they can make the best decision that suits them when it comes to taking an immediate callback, scheduling a later callback, or waiting on hold.

The success of any contact center is based upon the accuracy of the EWT estimation. For example, based upon the EWT estimation, the customer contact center can make decisions to efficiently route calls or to determine whether an offer for a callback should be placed. The accuracy of the EWT depends upon the accuracy of the conventional contact center’s statistics, since traditional statistical rules and heuristics are utilized to produce accurate EWTs.

However, the process of producing these EWTs can suffer from a variety of deficiencies. Our AI experts at Swampfox have worked relentlessly to get this calculation as accurate as possible in an effort to increase end customers willingness to participate in callback offerings. The more accurate we are with the EWT the more the end customer will trust their ability to receive the quality of care they expect.

To make our calculations more precise, we have aggregated various contact center operational data associated with time duration -as well as other proprietary metrics- to generate sets of data models that can continually improve as a best learning model to precisely predict accurate wait time. Some of these metrics include agent staffing, call arrival rate, call handling rate and seasonality. When this data is consumed and paired with our variety of machine learning algorithms, we can predict the EWT with a high degree of precision. To learn more detail about how we calculate EWT, please follow the link below to the patent.

LINK TO PATENT