“AI is selling like hotcakes” – Rowan Trollope
Enterprise subscriptions are the most important quarterly performance metric for cloud contact centre software provider Five9. And there’s been positive news. Subscription sales jumped 51% in Q4 2021. Large enterprises have been spending more. “AI is selling like cakes,” the company’s CEO at the time, Rowan Trollope, told ZDNet. What’s been attracting them? Artificial 
intelligence has become a tool ready for “primetime in call centers”, said Trollope. The former CEO said that a quarter of a trillion dollars was being
spent on experiences no one liked. AI, and more potently
Generative AI (GenAI), have the power to change this.
Five9 has launched its first ChatGPT-powered offerings (more on this later).
GenAI – early use cases are tilted towards agent assist
Businesses are experimenting with GenAI and early use cases are weighted towards agent assist. Let’s start with an example.
Dutch bank is ABN Amro has been trialling ChatGPT in its contact centre. The bank is now scaling it to 200 employees. Agents are often pre-occupied with taking notes during a conversation to write a summary afterwards. In the trial, ChatGPT has been taking the notes and agents have been reviewing them afterwards for accuracy. Employees now have time to focus on clients, instead of being distracted by time-zapping summarisation. Knowledge of our products is critical on the front line. ABN Amro is also using ChatGPT to summarise product pages.
Annerie Vreugdenhil, Chief Commercial Officer, Personal & Business Banking at the bank also shared some interesting insights on employee feedback. Fears associated 
with GenAI in the contact centre workplace are real. Initially, agents were worried about their jobs. Now they’re asking when the next GenAI feature will be deployed.
Developers are also happy with automatic coding, which makes them more productive too.
“We can’t let the AI run wild … There’s a limitation on what you can use it for” – Annerie Vreugdenhil
But, there was a red flag from Vreugdenhil: “We can’t let the AI run wild … There’s a limitation on what you can use it for.” The technology needs to be able to clearly explain its decisions.
Moreover, in any sector, GenAI requires masses of data. In a banking environment, where data is tied up in regulation, it’s too risky right now to use a tool like ChatGPT as a customer-facing technology.
Experiments right now, across industries, lean towards agent assist.
My job just got better
When will I be replaced by a robot? This is the existential question for workers staring at the automation precipice. In contact centres, like many workplaces, there are very real fears about the impact of AI on jobs – as the technology gets more sophisticated and interactions become near human-like. But evidence is all around us that we are thinking about this in the wrong way.
Customer service and support is a human experience. The evidence says that AI will not replace humans. It will give them superpowers to do their jobs better. Genesys research, for example, found that 47% of CX leaders say the top priority is investing in technology or connecting systems that improve the employee experience.
GenAI as a support tool will free agents up to focus on the conversation, more complex interactions, hone their skills and improve their product knowledge. The potential is there for agents to become highly-skilled, specialised contributors, which can only enhance the customer, and their own experience. As the ABN Amro example shows, early feedback from the bank’s employees is “my job just got better”.
Agent assist – practical use cases
Here are few practical use cases, paired with real-world examples and AI-powered innovations from some of our leading solutions providers.
Case summaries – as discussed earlier, automating case summaries saves agents time and effort. It also reduces Average Handle Time (AHT), a key KPI. A summarisation feature is already available in Five9’s Agent Assist. Prior AI versions could summarise conversations if they followed a specific structure – but it required model training and manual categorisation. AI Summary is an out-of-the-box solution that requires no training. It can generate summaries in seconds and publishes them in the CRM so the next person has context to answer the next call with personalised support. Richard Dumas, VP, GTM Strategy and Programs reports that saving even 1 minute from a 5 minute call means a 20% cost savings to the contact centre.
What can we learn from how brands are deploying summarisation technology?
Let me give you another example. Bouygues Telecom’s agents were struggling to capture information from the 8 million customer / agent conversations logged in its CRM. Moreover, they didn’t have time to automated transcripts from previous calls. IBM Consulting’s foundation models summarise calls, extract topics and updates CRM with instant actionable insights. Customers and agents have a better experience and the company has saved over $5m in operational improvements and a 30% reduction in pre- and-post call operations.
Automated knowledge articles and responses – GenAI can generate knowledge articles and human-like responses for agents to check and use. Five-month old Einstein GPT, the “world’s first” GenAI for CRM, helps service teams by automating knowledge articles from previous case notes and generates responses for agents, increasing customer satisfaction through faster, personalised interactions.
“GenAI can also understand intent, sentiment and emotion and generate empathetic human-like responses to help agents better serve customers’’ #callcenter #cx #AIgen #AI Share on X‘Experience orchestration’ – automating tasks and post-interaction follow ups a makes it easier and quicker for agents to serve customers. Genesys, driving innovation in AI-powered ‘experience orchestration’, has launched a new automation and task routing solution in Genesys Cloud CX. Many interactions require a follow up with another department. The technology automates workflows between contact centre employees and other departments for seamless end-to-end CX. Workflows can be customised. There is real-time visibility on every customer interaction to monitor status, completion and agent performance, and prioritise tasks in line with service-level agreements.
Extracting entities – “what’s the first line of the address” – agents ask for addresses line by line. How annoying is that? People don’t speak in that way – they want to give the whole address. Generative models can ask for a whole address, for example, and then decode the information into separate elements.
Agent training – the technology can analyse agent performance and customer conversations helping businesses to better understand, and plan training requirements. Again, there is a big caveat here on the technology explaining its decisions. There are HR implications here.
To conclude
Leaders are experimenting with using GenAI to enhance experience – for the customer and employee. They see customer service in contact centres differently. Their focus in not on call deflection and reducing waiting times. Their No 1 priority is investing in technology or connecting systems that improve the employee experience. These are the people that deliver great experiences. Self-serve is crucial. But do your customers really just want another chatbot?
The technology still needs to pass through validations. Early GenAI try-outs indicate it could finally be catching up with the promise – AI is enterprise-ready.
