Post-pandemic plans for digital transformation should include omnichannel communications, automation, and agent-assisted AI models.
TechRepublic’s Karen Roby talked with Colson Hillier, the CMO at Alorica, provider of Business Process and Customer Satisfaction Outsourcing solutions, about digital transformation projects in the post COVID-19 era. The following is an edited transcript of their conversation.
Karen Roby: Where should companies be focusing right now, and explain why omnichannel communications should be a priority.
Colson Hillier: I think that the idea of omnichannel is really about customer empowerment. Today, you think about the way that we interact with the brands that we love, and almost all of it involves multiple channels. I don’t sit and watch TV any longer just by itself. I’ve got my iPad up, and I’m emailing, or texting, or using social networks. All those channels have a different sort of role to play in the overall customer experience. It’s really important to get it right in terms of how you manage all of those channels in a way that’s consistent and delivers an outcome for a customer that doesn’t feel like it’s coming from six different places.
Our approach is to really look at that sort of opportunity on three different levels. The first is the application layer and making sure that you’re relevant and available at all of the different apps that are used by customers to understand your products, or get information, or make purchases. The second is an intelligence layer where you’ve got to apply a lot of the logic and intelligence on how to manage that customer interaction. And the bottom layer is knowledge management. So, you need to work all three of these in order to deliver the right experience, but knowledge management essentially takes all of the information that you need about product information, a customer’s profile, and puts it in a place that is accessible via the cloud and is well-indexed so that it can be used in those omnichannel experiences. The intelligence layer looks at all that information and takes the content of a query, whether that’s a social post, or a text, or a chat session or something like that, and it interprets that as the intent of the customer and delivers the knowledge that is needed to the client through that application.
If you can get those three things right, you can deliver an experience that feels very much aligned to where the customer wants to be met. There are different attitudes in social than you have when you’re SMS-ing versus when you’re on the phone with somebody. But the key is to make sure that you can manage that customer interaction across all of those channels in a way that’s consistent and sort of respects the interaction or the investment that the customer has made, either yesterday or earlier in the year, and give them an experience that’s personalized and sort of builds on your history with them, as opposed to feeling like it comes from six different places.
Karen Roby: Expand a bit on the importance of agent-assisted artificial intelligence (AI) models.
Colson Hillier: That core of being able to serve customers that have increasingly complex issues that they’re trying to resolve in an environment where our businesses are changing, our clients’ businesses are changing so rapidly, new product cycles, introductions, new policies and procedures. The pace of change is so great right now that it’s almost impossible to train an employee on how to handle every situation that they’re going to come across in a classroom training, what we would call just-in-case training.
A lot of where we focus our energy is on how do you take knowledge and use that at every stage of the agent life cycle to make sure that you have the right information for the client or the end user at the point of need? To do that, you need to be able to have your insights and information to resolve problems or provide customer experience readily available. It needs to be indexed in a way that it can be delivered in context into a conversation so that an agent is able to easily find and deliver that content to a customer.
We focused a lot on a couple of different things. One is that knowledge management layer–we mentioned that with omnichannel just a moment ago. The first way to deploy that is typically an agent-assist model. As you deploy an interface to an agent that’s handling a customer service interaction and they’re consistently utilizing this tool in order to access the content, and that content could come from the web, it could come from what we call tribal knowledge, so gathering all the intelligence of their peer groups, and supervisors, and things that typically were used on post-it notes or in notepads, and bringing that all together in a way that’s easily indexed and searched and delivered to the agent in a consumable format.
It can have a massive impact on the efficiency and the consistency of the delivery of a customer-service experience. For example, when we looked at one of our clients, we started by evaluating all the calls they were taking. And we found that over 45% of the time an agent was on a call with a client, it was dead air time. And that time was spent researching and looking things up, it was asking a neighbor or a supervisor for some support, or generally processing things that were very low value to the customer interaction and really driving inefficiency in that interaction. So, when we deploy a knowledge management solution, the agent has access to that information in real time and is able to get those responses delivered much more quickly. We see average handle time reduced by about 15%, we see what we call first-call resolution, or the ability to resolve a customer’s issue the first time that they call in, increased by seven to nine points.
These are really important metrics to the efficiency of how you deliver a customer experience through the call centers that we operate. The other important byproduct of having a platform like that is that it deploys AI and machine learning in a continued-improvement process. As I start to get questions, and I’m interrogating this knowledge-management platform to get the answer, if that’s the correct answer, I know that, and I sort of signal back on the affirmative. But if it’s not, I’m able to go back and have our knowledge engineers go back and understand why that was the case and tune the model so that it gets more effective. As you do this hundreds of thousands of times, you get to identify things that are repeatable and that you know you have high confidence in being able to deliver the response against. Those are the best candidates to take and migrate into something that’s fully automated, like a bot experience where humans are no longer required to deal with some of the types of questions that come up in our call center environments.
Karen Roby: What are your thoughts on automation?
Colson Hillier: There’s so many companies that have grown up, particularly larger companies, where processes that worked well at small scale, or when you had a very streamlined decision-making flow, or a limited number of systems that needed to be touched, worked just fine. But as you go through growth in your business and add new departments that are part of a chain or new systems that need to be updated and kept in sync with others, we found that process automation is an awesome way to sort of thread together all of the disparities in a large ecosystem that needs to be managed in order to deliver an outcome.
Whether it’s something like a back-office process automation where you’ve got an intake of a document that needs to be updated across four or five different systems and done consistently with high degrees of accuracy, or if it’s a process that is creating a lot of fallout or negative net promoter scores that you can address with systematic delivery, the process of understanding the end-to-end flow of information as it moves through a company, identifying where there are significant areas of fallout or inconsistency, or even very high resource allocation, like labor allocation, helps us to identify where there are opportunities to, again, deliver efficiency, consistency, and ultimately a better customer experience by solving on speed and accuracy.
This article originally appeared in TechRepublic: https://tek.io/3ej2moB