How to deliver great phone service with AI agents

More and more companies are implementing AI-powered voice agents to develop self-service over the phone. Available immediately, 24 hours a day and insensitive to call spikes, these tools can deeply improve customer service performance. However, they raise questions in terms of CX: how to ensure that callers will reach their goals easily with a voice agent? What will their experience look like? Discover the 7 key rules to deliver the best Caller Experience for your customers.

According to a recent NICE inContact’s “CX Transformation Benchmark”, 39% of customer interactions with companies now take place in self-service. Customer relations are no longer only about people: virtual assistants are being developed on various channels to help customers solve their problems in complete autonomy.

AI-powered voice agents are expanding over the phone

The telephone channel remains the preferred channel for customers to contact a company (two-thirds of interactions). Yet it is probably the hardest to handle: massive sollicitations, answers to provide in real time, risk of call spikes...

This explains the emergence of voice agents, virtual assistants capable of offering self-service over the phone. These AI-powered agents improve the caller experience, thanks to their 24/7 immediate availability.

They also allow companies to drastically increase their performance by handling repetitive calls, allowing human agents to focus on high value-added interactions.

How do AI voice agents impact customer experience?

Voice agents can ensure call qualification and routing to the right human agent, or automate end-to-end use cases such as appointment scheduling or request tracking.

However, this new technology can also generate worry, especially at the CX level. How to make sure that implementing a voice agent is relevant to the customer journey? And what best practices should be followed to facilitate the development of self-service over the telephone in your company?

7 rules to offer the best Caller Experience with AI voice agents

Creating and delivering a good experience over the phone when talking with a voice agent is a real discipline. At calldesk, several members of our R&D team are working on the caller experience. Here is an overview of 7 key rules to follow to deliver the best caller experience with a voice agent.

Rule #1 - Choose the right use case before deploying the voice agent

Voice agets must respond to a clearly identified business problem. This tool must allow you to achieve your goal, whether it is to reduce the pressure on your contact center, to enrich your caller experience or to improve your team’s work environment.

Voice agents are particularly suited for handling massive amounts of requests that generate several thousand calls per month, and for which human intervention provides little to no added value. The voice agents can also be deployed tactically, to handle calls that the contact center is unable to absorb (e.g. during call spikes, or to offer evening and weekend availability).

These use cases can be identified using a diagnosis of call flows, an audit approach that prioritizes the calls to be automated in order to quickly generate ROI.

Rule #2 - Reassure the caller at the beginning of the conversation

Your AI voice agents are here to reinforce your human agent teams, not to replace them. The alliance between human and AI-powered agents is the most effective way to maximize customer satisfaction.

From the start of the conversation, clearly indicate to your callers that, if necessary, a human agent will take over to process their request, either by transferring it immediately or by scheduling an appointment later.

Rule #3 - Avoid the uncanny valley

The more a robot tries to imitate human behavior, the more its quirks and mistakes will seem disturbing and cause discomfort. This phenomenon has been called the "uncanny valley".

Don't try to humanize your voice agents at all costs, for example by giving them a first name or by using onomatopoeias. Enhancing the experience does not mean faking human behavior.

On the contrary, it can irritate or mislead your callers, and can even confuse some fragile customers, such as elderly people who may think they are talking to a human - it happens more than you could imagine!

Rule #4 - Engage the caller in using the voice agent

Some customers will tell you that they don't like talking to robots. In fact, the problem is often hidden: customers reject the technology because they don't understand why it was implemented.

By explaining at the beginning of the call the value that the voice agent will bring to the customer, and the steps needed to achieve this, the acceptance rate increases drastically. Your customers' effort is compensated by the value they get from the conversation.

For example, you can explain that the voice agent will allow your customer to be taken care of without a queue time, or that he will get the information he needs to start processing his case, etc.

Rule #5 - Guide the caller towards problem resolution

To complete the conversation and allow the callers to solve his problem, you need to guide him in the right direction. Therefore, the voice agent must take control of the conversation, to avoid confusing the customer about what he should say. With a good voice agent, the customer is carried from the start of the conversation to the resolution of the problem.

Hence, the voice agent can act as a human agent would: for example, it can confirm the information retrieved during the call, by calling the caller by his name ("All right [first name], how can I help you today?").

He can also reformulate the caller's answer to ensure that he understood well.

Rule #6 - Avoid repetitions

Choosing the right words is important! Indeed, repeating exactly the same sentences at each turn of the dialogue may highlight the robotic side of your voice agents.

On the contrary, diversify the vocabulary used, know how to give the right information at the right time, and don't hesitate to reformulate the questions asked by the robot in case of misunderstanding. Just like a human would do spontaneously! The goal of your voice agents is to become fluent in human language.

Rule #7 - Choose the right technology for your voice agent

There is a direct link between voice agents' technological performance and customer satisfaction. For example, offering a good comprehension rate, despite background noise or customer accents, is fundamental to allow the conversation to proceed naturally.

Similarly, latency is important, as it directly influences the voice agents' response time. If your callers have to wait several seconds after each answer, the conversation will become completely unnatural.

Finally, make sure to choose a solution suited to the voice channel requirements. For example, when we answer a question orally, we often deliver several pieces of information at once: "no, I no longer live in London, I moved to Birmingham". In this situation, the voice agents have to understand both that the customer's address needs to be updated and that the person now lives in Birmingham.

In conclusion, the deployment of a voice agent project must address a clear business need, and take into account the caller experience. By applying a clear discipline, the Caller Experience, to the creation of conversations, you will get your customers' buy-in and you will combine the development of self-service with increased customer satisfaction.

To discover more about AI-powered voice agents, visit calldesk.ai.