How to get the best from AI-powered engagements

The UK Customer Satisfaction Index has been heading south for the last five surveys, recently reaching its lowest level since 2015. The biggest drops were in quality of interactions, standards of complaint handling, and the extent to which companies build (or don’t build) experiences around customer needs.

So, what can we do to meet and beat customer expectations in this new decade? Well, the digital channel creating the most buzz is artificial intelligence (AI). So, looking beyond the hype, what practical steps can we take? And what’s the best way to go about it?

Getting under the skin of conversations

Customers want to feel an emotional connection with a brand, not just sterile automated transactions. Trust is founded on empathy and AI can help here by detecting hidden human emotion in phone and chat conversations. Many organisations have already figured this out and are deploying AI as a customer experience augmenter to deepen relationships.

Even agents, once fearful of the impact that AI might have on their jobs, have become more receptive. In a poll of 7,000 service staff, 60% felt AI would make their job simpler and help them work more effectively. Meanwhile, the proportion concerned that technology may replace them was down at a mere 15%.

Where to deploy AI for C-SAT and NPS uplift

Start by mapping journeys and, where it makes sense, consider inserting voice and chatbots to reduce customer effort for basic requests and information. That way enquiries get solved faster, while live agents are released to better handle more complex issues.

Think about deploying AI internally. For example, to uncover learning at scale from myriad historic contacts. Or track agent performance and flag training needs. That’s what DNB did. The Norwegian bank identified over 200 different competencies and skills involved in resolving customer inquiries. Then, using learning algorithms, it pinpointed opportunities for increasing first contact resolution through targeted coaching.

Those are two great ways to road test AI with relatively small risk and big potential for boosting customer satisfaction and net promoter scores.

Using AI to boost revenue

Outside the contact centre, marketing departments are using artificial intelligence and machine learning to convert more website visitors into customers. It’s all about proactively engaging at exactly the right moment.

For example, by identifying when customers are not going down the right path, or when they get lost. Or, crucially, rapidly heading off shopping carts on the verge of abandonment. AI is perfect for this.

Most popular AI deployment models

Given the growing maturity of AI-powered chatbots and virtual assistants, Gartner believes 15% of all customer service interactions will be handled solely by AI by 2021. That’s a 400% increase from 2017. And this will be largely powered by the cloud, with a study by Deloitte suggesting 70% of companies will be taking that route. It concludes that in the coming year the penetration rate of enterprise software with built-in artificial intelligence, along with cloud-based AI development services, will reach an estimated 87% and 83% respectively.

With its open architecture and APIs Genesys Cloud is purpose-built for deploying blended AI and integrating with third-party AI solutions. For more strategic approaches to CX in the contact centre, view the report How to deliver customer experience in the 2020s.