No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

AI Comes to the Contact Center

After a long period of stagnation, the contact center industry is experiencing a renaissance of innovation. Contact centers in the cloud are providing enterprises with new options and economics for infrastructure. And omnichannel platforms are allowing brands to get closer to their current and prospective customers through a variety of touch points, including voice, chat, video and social media channels.

Once the customer establishes contact, focus turns to optimizing the outcome for both parties. And that is perhaps where the most revolutionary innovation is occurring: Artificial intelligence is increasingly being deployed to establish the most productive and mutually beneficial path for each customer conversation.

Historically, the automated call distribution (ACD) system would route a call to the first available agent (first in, first out -- or FIFO). That approach was augmented by skills-based routing (SBR), in which calls would flow based on organizational logic such as subject matter or issue type. So if the caller indicated a need for help with billing, or with setting up voicemail on his or her smartphone, the system would identify the list of agents best trained or experienced in handling such a call and then put the caller in queue for the next available agent from the specialized list.

These traditional approaches rely on relatively straightforward, tangible and objective decision criteria, and have only marginally been updated in the past two decades. More recently, advances in artificial intelligence and big data have enabled routing based on more complex bases. Chief among these is personality, or behavior. It's common sense that a caller and an agent who establish a rapport are more likely to achieve success together. Now science has caught up with common sense in the form of enterprise behavior matching (EBM).

EBM represents an alignment of three sets of variables:

The first two components of the match are obvious. Given the diversity of industries in which companies operate contact centers (consumer products, financial services, insurance, telecommunications) as well as the types of contact centers (customer service, telesales, collections), the organization through which the customer and agent connect heavily influences the outcome as well. In other words, to be as effective as possible, an EBM model must be tailored to the enterprise and environment in which it is to be deployed.

One driver of EBM success is ease of integration with existing contact center infrastructure. Telephony, chat and email platforms, customer relationship, and workforce management systems are often well entrenched in the enterprise, and new technology needs to be capable of fitting into or overlaying that infrastructure with minimal disruption in order to be accepted and be effective. EBM should add no delay following existing IVR-ACD-SBR routing. Of course, the term infrastructure should really be understood to include business operations and the culture of the affected contact center. This means managers should perceive that EBM is immediately adding value, and agents should be comfortable that it is improving their workflow and results.

The rules underlying traditional forms of routing such as ACD or SBR generally change infrequently. In contrast, the consumer, agent and organizational behaviors driving EBM are infinitely dynamic, and EBM systems therefore must learn and adapt continuously.

Millions of records of new data must be analyzed on a daily basis, and models and matching algorithms must be adjusted just as quickly. In addition to updated public demographic data on consumers and profiles of new agents, call outcomes from the day need to be incorporated into the analysis -- And the analysis needs to be exhaustive. For example, the introduction of a new advertising campaign might dramatically alter customer behavior and skew call outcomes compared to previous experience.

As much as enterprises deploy and upgrade contact center technology to serve their end customers more responsively and efficiently, one constant remains: the customer desire for a human at the other end of the connection. While artificial intelligence has been used to simulate human response, its real power lies in optimizing the interactions between two humans, and enterprise behavior matching is one such application that is already achieving results in contact centers today.