Your Genesys Blog Subscription has been confirmed!
Please add firstname.lastname@example.org to your safe sender list to ensure you receive the weekly blog notifications.
Subscribe to our free newsletter and get blog updates in your inbox
Don't Show This Again.
There’s a new dawn in the role of the contact center agent. Over the last decade, the maturity of conversational artificial intelligence (AI) has seen a massive shift of customer care and support to self-service. Self-service automation tools and IVR like bots and knowledge centers can easily handle rudimentary tasks, such as checking an order status or troubleshooting an error code. Consumers have fully adapted to this self-serving paradigm and will look to use those options where they can.
Today, consumers interacting with human agents have much higher expectations and demands than they did a decade ago. For the agent, there are no “easy” calls anymore — bots have taken care of those. If a customer is speaking to an agent, it’s because they couldn’t get their issue resolved via self-service because it was either too complex or required a significant amount of human understanding and empathy. Combine this with the new work-from-home norm, where you don’t have a wise desk buddy or supervisor sitting by you, ready to jump in with help.
In addition to this, everyone is facing budgetary pressures. Doing more with less means you want to solve issues in the most cost-effective way you can. For example, an appliance shop would rather help a customer troubleshoot their own appliance as opposed to sending out a repair technician, especially within the warranty window.
This has made the already difficult job of customer service even more difficult. Today, contact center agents are facing way more complex queries with higher customer expectations and more pressure to resolve the call — all while working remotely without the benefit of turning the call over to a more experienced peer. And this is leading to increased agent burnout and high turnover rates.
AI can bridge the gap created by these new work practices so agents can meet escalating demands.
Let’s set the scene. Malcolm is 3-4 months into his new agent role. He’s taken the mandatory training, has received a laptop and headset, and is working from home. A customer who needs to speak to an agent because of a complex query is routed to Malcolm. The customer is already a little frustrated and impatient because she started on a self-service channel in chat but couldn’t find a resolution. Speaking to an agent is her last line of inquiry.
The customer explains her issue, which is complex. Malcolm can’t put her on hold and ask his supervisor for help. The supervisor isn’t immediately available, and the clock is ticking — every second matters. According to the recent Genesys “State of Customer Experience,” report, 33% of consumers say they’ve stopped using a company after a single negative service interaction in the past year. So, maintaining the optimal average handle time is critical to create an exceptional customer experience and build loyalty.
The agent has access to the knowledge base, so he switches over to another screen and starts looking. The customer is on hold and time is ticking while Malcolm scans knowledge articles for an answer. The customer is now more frustrated — and the agent is stressed. It’s an unpleasant experience for everyone.
This is where AI can help. Contact centers can use AI-enabled knowledge to listen to the customer, identify a complex query, and find and surface up the right answer to the agent in real time. This means having a knowledge base that’s optimized for semantic search and uses additional AI to find and simplify the information.
When AI technology equips agents with the answer, there’s no need for customers to wait on hold. Information is retrieved as the conversation is happening and, with a click or two, the inquiry is resolved. And this technology isn’t just for digital interactions; agents have access to real-time transcriptions from phone conversations and benefit from real-time contextual information.
AI can also drive the ability to improve the customer experience through predictive engagement. It can listen to a customer’s behavior and then automatically calculate a segment, or a predicted outcome, based on behavior patterns. This data is often used to trigger an offer for cross-sell or up-sell or conversation. For example, AI-based predictive engagement can target a conversation about mobile devices to a bot for a customer who belongs to an identified mobile segment — and who is more likely to purchase.
The same capability can be used to show the agent the entire customer journey across multiple interactions, including any blockers. Giving the call center agent context about the entire customer journey can improve agent efficiency, improve customer satisfaction (CSAT) and, ultimately, help drive a better outcome. The agent has insight into why the customer is contacting the company and if they’ve encountered any issues in the journey.
AI in the contact center can also improve the agent-and-customer connection before the interaction even begins. And while using artificial intelligence to optimize how interactions are routed isn’t a new idea, it’s been traditionally difficult.
Previous call center software required an army of data scientists who would cull through existing interaction data, build models, test models, and then deploy and measure them. However, Genesys AI technology enables this to be a three-click process:
This also creates multiple built-in reports that show how predictive routing is doing and if it’s achieving set KPIs, including a model viewer that shows what interaction or customer characteristics have the biggest impact on the target KPI.
Generative AI can — and does — play a massive role in a contact center agent’s day-to-day life in summarizing customer interactions. This can be a time-consuming and error-prone task that’s commonly referred to as an “After Call Work” in the contact center. There are specific and targeted ways this job needs to be done. For example, an agent does not need generative AI to summarize in the form of a limerick.
A caller had a coffee machine woe
And contacted Support to know
The agent named Grace
Found a recall was in place
And a replacement machine did bestow
This example was generated using an open-source AI with 175 billion parameters. While it’s fun, it’s not necessary. However, contact center employees can use generative AI to summarize and capture the key turns, intent and outcomes of the conversation. It’s also worth noting that these can domain-specific.
Reason: Coffee machine issue
Customer Intent: Resolve issue
Outcome: Replacement machine dispatched
Final Customer Sentiment: Positive
Summary: Customer informed agent of issue with their machine. The agent asked for the part number. Agent found that there is a recall on part number ST145. Agent arranged for replacement to be delivered to customer address to collect old machine.
This summary was generated with a much smaller 780 million parameter Large Language Model (LLM). And the model was trained on contact center use cases.
To deliver the personalized end-to-end experiences that customers want, contact center employees need time, context, and quick and easy access to data and information. This can be accomplished with three things:
Watch the following video to learn more about how Genesys Cloud Agent Assist empowers your organization to orchestrate seamless end-to-end customer experiences. And then try Agent Assist by downloading it from the Genesys AppFoundry® Marketplace.
Subscribe to our free newsletter and get blog updates in your inbox.