Byron Copley at Five9 explains artificial intelligence-driven customer experience in the contact centre space.
AI Must Enhance the Customer Experience
Ideally, an AI-driven customer experience quickly assesses and resolves your customers’ concerns and keeps them engaged with proper next steps.
It’s becoming clearer that AI-driven technology, which includes Chatbots, Conversational Agents (CA), Interactive Voice Recognition (IVR), and Intelligent Virtual Agents (IVA), can provide an excellent customer experience.
In this blog, we’ll examine three points:
- How AI drives improved customer satisfaction in the contact centre
- Measuring AI’s effectiveness in the process
- How AI needs to improve in the contact centre environment
We conclude that AI in the contact centre space has an important and ever-expanding role in shaping the customer experience, but it’s not yet ready to supplant all human agents.
1% of Surveyed Customers Answered “Absolutely”
Your customers expect a brief, seamless, friendly, frictionless, and connected experience when they interact with a contact centre.
“Resolving a complex situation with an enterprise-sized contact centre operation” is not on their bucket lists, so it’s already an inconvenience for customers to embark on a journey through the contact centre matrix.
In fact, according to Talkdesk, 43% of customers answered “No” to the question: “Do customer support contact centres always provide excellent service?” And 11% said, “Absolutely…not.” Only 18% answered “Yes.”
And a mere 1% answered “Absolutely.”
The Talkdesk study clarifies: “The majority of customers surveyed are dissatisfied with the service they receive from contact centres.”
So, it’s clear that customers are highly likely to expect an unsatisfactory experience before they even interact with a contact centre. And yet, the same study also concludes that customers’ top three priorities are these:
- Problem solved quickly
- Personal interaction with an agent
- Speak with a skilled agent
Overcoming this conflict between what customers expect and what customers experience poses a significant challenge for contact centres.
What Is the Best AI-Driven Contact Centre Customer Experience?
Customers get quick resolutions to their problems at less expense to contact centres, while human agents are available to solve more complex customer concerns. Everyone benefits from this type of AI-enhanced customer service.
When AI does the heavy lifting, it’s called Self-Help Level 0 resolution. This best-case scenario resolves a customer issue without involving a service agent.
This outcome is easier said than done. When it happens, it’s called “containment”, as in the customer was “contained” within AI technology. In your customers’ minds, the intrinsic nature of contact centre automation sets the expectation of quick, favorable resolution.
Immediate technology; immediate results. No detours, delays, or distractions. A straight, continuous line from the problem to the solution. More like a customer dash than a journey.
After all, according to Statista, 42 percent of contact centre customers still prefer the phone as the main method of communication.
However, that does not mean customers necessarily always want to talk to a live agent. Self-Help Level 0 is the best case for businesses, since, according to HDI, transferring to Level 1, which is usually the service desk, costs a business 10x more than Level 0.
Even when it’s necessary to involve a human agent, your customers expect quick, seamless handoffs with agents who are informed of their situations with the right balance of personalization.
Like we said—easier said than done. Yet it’s what contact centres everywhere should strive to achieve, and many are gaining effectiveness at resolving basic and even more involved customer concerns with AI.
AI Needs “Training” Before It Can Respond to Customer Requests
With the present sophistication of AI technology (or relative lack thereof), the ability to provide a pristine data set helps ensure the effectiveness of AI and “train” it to respond correctly to your customers’ requests.
It can provide enormous quantities of contextual data in real-time that continually improves the intuitiveness and insight of AI technology for the contact centre.
These are some of the Key Performance Indicators (KPIs) that are critical to measure the effectiveness of AI technology:
- Self-Service Engagement: The percentage of users who attempt to engage with AI, as opposed to those who disconnect or immediately “zero out” to speak to an agent
- Task Completion Rate: The percentage of tasks successfully completed within AI channels
- Cost Per Transaction: Cost of all contacts handled in AI channels
- Containment: The percentage of interactions that begin and end in the AI system and do not require any agent intervention
A platform integrates all these metrics, and many others, into one data source and visualizes them on dashboards for review and analysis.
This enables contact centre managers to make real-time adjustments to their processes to increase CX and CSAT.
IVAs and Chatbots Need to Be “Human” in Nature
Aceyus, the company Five9 purchased in August 2023, has offered content worth reviewing in past blogs about chatbots, conversational agents as well as IVR and IVA.
Maybe what we haven’t discussed to date is the present state of contact centre AI technology. Ultimately, IVA and chatbots will take on a larger share of traffic that human agents currently absorb.
Realistically, though, most of this contact centre AI-based technology is, in many ways, still in the “toddler” stage.
While some organizations have developed AI that handles complex tasks, a majority of AI today efficiently completes simple tasks, including call routing, bank balances, accepting payments, password resets—tasks that free up agents to help solve more complex customer scenarios.
However, AI can create a new portrait that looks like Rembrandt himself painted it. Can it do more than answer a few basic questions? Can it even conduct a simple conversation?
Five9 has made great strides in this area, with more organic, responsive conversations as AI “learns” to “behave” and “respond” more like a human being.
As stated “A Virtual Agent is about 10% the cost of a live agent, available 24/7, speaks 120 different languages, provides 100% accurate information, and can effectively read customer sentiment so calls can be escalated to Tier 2 support when necessary.”
Study: AI Lacks Social Presence
Studies suggest that people inject a social element into a conversation with a CA, the “typing” version of an IVA.
Your customers know they are interacting with computer software, yet they expect a reciprocal “human” reaction from AI.
However, studies like this one from March 2020 conclude that CAs are not mature enough to reciprocate on a “human-enough” level.
They lack the social presence, aka “skills”, of customers and agents. As a result, these positive-goal-seeking mechanisms guide customers down blind-alley outcomes rather than prompt, satisfactory solutions.
To quote this study: “Though cost- and time-saving opportunities triggered a widespread implementation of AI-based chatbots, they still frequently fail to meet customer expectations (italics ours), potentially resulting in users being less inclined to comply with requests made by the chatbot.”
The Key to Machine Learning: Data
A long customer journey risks the exact opposite outcome that customers and businesses desire. But there’s a conundrum at play here: AI must “gain experience” to “learn” and become more “human” through machine learning.
And the key to machine learning is data—and lots of it. Machine learning shapes and adjusts critical algorithms that refine predictable patterns and responses to keep customers solve their specific challenges.
So, as AI technology matures, contact centres should keep pace with rising customer expectations.This blog post has been re-published by kind permission of Five9 – View the Original Article
For more information about Five9 - visit the Five9 Website
Call Centre Helper is not responsible for the content of these guest blog posts. The opinions expressed in this article are those of the author, and do not necessarily reflect those of Call Centre Helper.