AI bots for customer experience: trends, insights, and examples

The hype surrounding AI-based voice and chatbots is evident, but do they deliver? Most still perform only extremely basic tasks and often mirror the poor practices of traditional IVRs. Customers may be open to the idea, but only 30% believe that chatbots and virtual assistants make it easier to address their service issues.

The things customers say bots are good at – replying quickly, helping them outside normal business hours, being friendly – are not as important as what 60% say bots fail at: understanding them well. Any company that underestimates the value of AI in this area will fall behind in 2023; however, reevaluation is necessary. 

How can you implement AI bots in your company, and what will they be able to do for you? Here’s how Avaya expects things to shake out: 

1. Companies will take a more realistic approach to bots with a strong focus on improving dialogues and designing superb services behind conversational user interfaces

Customer expectations are drastically changing with the proliferation of AI-enabled speech devices such as Alexa, Google Home, and Siri and the introduction of new applications like ChatGPT. We will see a focus on making customer interactions easier to understand, resulting in less repetition of information and disconnects to create better bot experiences. Integration with cognitive intelligence (context-sensitive knowledge management, predictive analytics, and similar) will be key for doing so.

Don’t expect to see night-and-day differences but rather a turning point in the relationship between machines and humans. We’ll see the continued shift away from button-based chatbots to conversational virtual agents that can handle more complex interactions with a sophisticated reasoning engine and integrated back-end systems. 

Companies would be wise to start with small-scale, attainable applications (ex: having their virtual agent successfully “talk” customers through the steps required for a given task or process such as initiating a password reset, filing a support ticket, or making a reservation). Then, build from there to start using conversational AI in more advanced ways such as instantly capturing and analyzing conversations to initiate action or having bots read between the lines to understand what customers want and feel. 

For now, it’s “slow and steady wins the race.”

• Handing off interactions from live agents to bots (not just from bot to agent) for process completion will become an increasingly interesting option

Most bot-initiated interactions result in live agent transfers. Could we see an increase in reverse handoffs where customers go from agent to bot? Yes, especially as processes near completion so that agents can free up more time. 

Here’s an example of how this could work: A customer calls his car insurance company after being in a minor fender bender. The agent walks the customer through the steps that need to be taken (how to file a claim, where to upload pictures of the scene, and how to handle all other needs) and answers any questions along the way. 

It’s at this point the agent offers the customer the option to be transferred to a digital experience facilitated by the company’s AI chatbot. The customer accepts, the call ends, and the bot steps in via text/SMS. The bot can send links to knowledge-based articles, embed “how-to” videos directly into text messages, help the customer navigate the company’s mobile app, and answer any questions in a natural, human-like, back-and-forth conversation. 

The customer can opt out of this digital experience at any point and be elevated back to a live agent with all contextual information. Avaya expects more organizations will consider or facilitate agent-to-bot interactions in 2023 to improve costs and productivity while maintaining, if not increasing, CSAT. 

• This will be the year when CX no longer depends on contact center platform features but on the platform’s ability to collect, process, and react to data

Expect to see a surge in interest for communication analytics as companies seek to convert data hidden in customers’ conversations into actionable insights that can be used to make strategic, tactical, and operational decisions. For example, using AI-powered speech analytics to: 

• Put together accurate profiles of individual customers based on their interests and attributes so they can be strategically used in future conversations (both bot and human).
• Generate key insights about a customer group and proactively communicate (via bot or automated notifications).
• Help your bots understand what customers are requesting (even if phrased in an odd or unexpected way) to recognize their aim in starting a conversation. 

An AI-powered Experience Platform, built on open API architecture, provides the predictive, cloud-based AI capabilities needed to gain deeper customer insights, personalize interactions, and reduce operating costs while reaping the benefits of continuous improvements over time. The platform seamlessly integrates with existing infrastructure, preserving current investments, while empowering organizations to find valuable data immediately and start managing and leveraging that data more effectively. 

An effortless way to get started with AI bots is by using a pre-built VA solution. These solutions are the epitome of “innovation without disruption.” They can be managed and customized from a simple dashboard without needing help from a developer team and can be used for simple to very sophisticated purposes. It’s likely we’ll see an increase in the adoption of these solutions over the next 12 months. 

How do you see AI bots continuing to evolve for customer experience?