Businesses are employing hyperautomation, which is a structured approach to quickly automate as many processes and workflows as possible. In customer service, the goal of hyperautomation is to ultimately improve customer experience (CX), increase operational efficiencies, and reduce costs.
The contact center is core to the success of the business, and it’s undergoing the transformation brought on by hyperautomation. In a recent session at the Gartner 2021 Application Innovation & Business Solutions Summit, Magnus Revang, a VP analyst at Gartner, discussed how to build a toolkit of technologies to succeed with hyperautomation.
There are three important considerations for any automation project:
- “Deciding what to automate”
- “Picking the right tool for the job”
- “Sustaining automation”
One of the automation tools that contact centers will surely use is conversational AI. Using conversational AI, contact centers can quickly resolve Tier-1 issues, eliminate hold times, and free human agents to focus on more complex issues.
Here’s what contact centers need to know about building their toolkit and what to consider when incorporating conversational AI into that kit.
What should you automate?
There are two sets of concepts that are important to understand when determining what you should automate. The first is automation potential versus automation grade. According to Magnus Revang, automation potential is the “[a]mount of work that can be automated with available technology.” Automation grade is “[h]ow much of your automation potential have you realized.”
The next set of concepts is complexity versus sophistication. Complexity is the sheer amount of work and effort that goes into automating a particular piece of work. Sophistication is the level of capability that you need to do it. Are you automating something that’s highly complex, highly sophisticated, or both?
Revang says, “Tools add capabilities, but only increase your automation potential.” When you add more tools to your toolkit, you’re only increasing your automation potential to cover more and more sophisticated tasks, processes, interactions, and decisions. You’re actually not automating because automation is not automatic.
Which tool is right for the job?
There’s an enormous amount of automation tools available to you. These categories of tools are all expanding and increasingly overlapping with one another. For example, conversational AI platforms also touch AI and machine learning, RPA, AI knowledge management, and more categories. As a result, you may not need tools for each category. It is important to evaluate your current tools and how their capabilities have expanded over time.
At the same time, there are new types of tools emerging, including hyperautomation platforms and AI decision systems.
With such a crowded landscape, “you have to determine what you’re automating, and the sustainability of the tool.” Are you automating a task, process, interaction, or decision? Each has different tools that are best suited for the job.
From there, you need to determine the level of skill and effort needed to realize the automation potential. For conversational AI, the skill and effort levels required ranges from low to high. Managed services bring expertise and resources to the table, so you just need low or medium skills and effort. APIs, such as Google Dialogflow or Amazon Lex, require high skill and effort, since you’re building your own AI.
Take a look at what skills you have in your organization today. “The more available the skill set is, the more effort is potentially available to you.” If you’re able to train more people to build or QA the AI, you’re more equipped to succeed with building your own conversational AI.
Is this automation sustainable?
When deciding on the tools you should invest in, the last thing to ask yourself is “How adaptable to changing conditions does your automation effort need to be?” Automation projects often get stuck in the list of IT requests and aren’t acted on unless they’re the number one priority. The automation stays stagnant, despite everything else in your business changing. To avoid this, you need to consider how adaptable your tools need to be.
Think about how frequently the automation needs to change the level of complexity and sophistication of these changes. This will determine whether you need a dedicated team to manage the automation and what that team will look like.
The field of AI is growing rapidly. In order to keep up with customers’ expectations and your competitors, your conversational AI models need to be frequently updated and retrained. While changing what the AI says to a customer is a low skill and effort task, training the AI models isn’t. Do you already have a team to make these changes, or can you get a team to do so?
When deciding what type of conversational AI tool to use, make sure you choose the sustainable option. If you’re a contact center that has few engineering resources, partnering with a managed service will ensure the sustainability and adaptability of the conversational AI. Building conversational AI yourself means you’ll need to have a dedicated team to maintain it over time.
Hyperautomation can deliver amazing results to enterprises and contact centers, and conversational AI is one of the main tools that’s being used by contact centers to automate customer service interactions. If you’re ready to add conversational AI to your own automation toolkit, learn how Replicant Voice can help you automate customer service calls.
Gartner, ‘Application Innovation & Business Solutions Summit – Americas’, Presentation (The Software Engineers Toolkit for HyperAutomation), Magnus Revang, May 26-27, 2021