Ten years after we were told that Software is Eating the World, we’re now well aware that The Cloud is Eating the World. Hosted contact center software offerings continue to grow in number and sophistication, while legacy players rush to transform their products accordingly. CRM has long been in the cloud. COVID sent agents to work remotely, and the current job market will keep them there. We get it. There is little sense for companies to host these business critical applications themselves, for reasons we don’t need to recount here.
And then there’s AI. Everyone wants it, needs it, and claims they have it. So in a world where every product description ends in “aaS” from SaaS to CCaaS, it only makes sense that AI should follow suit.
Not. So. Fast.
As much as we’d love to tick a few boxes, drag and drop a few user journeys, and see AHT fall and CSAT rise, we all know it’s not that simple. While many of us want to see AI scale as quickly as other hosted applications, the simple truth is that it just doesn’t work that way.
The way humans describe their customer service needs vary widely from business to business, and they vary massively from person to person. This means that while AI technology is repeatable, the training data that these AI models rely on isn’t. In other words, AI needs to be trained to understand the nuances of an individual business’ customer experience and journey.
If you, dear reader, have made it this far and I’ve made you scared, please don’t be. There are many incredible case studies of AI positively transforming the contact center. The key is not to rush to implement AI to automate, and instead first implement AI to understand.
By first training AI to understand the customer experience in detail, contact center leaders will have the insights on the conversations that can be automated well and those that are best left to human agents. This first step is easy, requires little technical lift, and has the added benefit of providing insights to training, marketing and product teams. And once deployed for automation, effective AI investment drives net revenue savings, effectively financing itself.
In the same way it’s a faux pas for engineers to test their new software in production, it’s just as ill advised to train an AI in the wild with actual customers who need their issues solved. So for those looking to embrace AI, you don’t need to go buy servers, but you should consult some experts in the space, and take some time to tailor your strategy and integration before setting it live in the wild.
I talked in greater detail with David Hadobas and Vince Lynch. If you’d like to learn more you can watch the video. If you still have questions, feel free to find me in the CCNG Slack group or at owen@iv.ai
Owen McGrath is an active CCNG member based in Boulder Colorado and the Head of Sales for IV.AI.
