One of the differences between conversational AI and, say, your website or most other channels that don’t rely on natural language is that…
When you have a conversation, your expectations are raised.
I’ll give you an example: if I have a billing issue with my streaming provider, I’ll go to the website and I’m likely to find generic information because the information on the website is confined to whatever the company decides to put on their website. It’s not going to be specific to me and my situation.
Compare that with a conversation that I would have if I was to phone the contact center. I expect an answer specific to me.
I expect to know whether it was a billing issue because I ran out of money, whether my card expired, whether my free trial expired. Why did my specific account stop working? And I would expect the person who I’m speaking to, to be able to answer me that question. They have access to the systems that have the information buried inside them, after all.
And so when you start thinking about conversational automation, it’s not really a question of conversational AI technology: conversational AI platforms, conversation design technique etc. It’s more of a question of your internal architecture, your technology infrastructure and creating the environment and the technical infrastructure and capability needed to be able to actually have those useful conversations.
Now, you might not start with those integrations. You might start with more information-based, simpler transactions. But when you do start thinking about your conversational AI strategy, one of the first things you need to do is involve the people who own and manage the technology because it might take a little bit longer than you think to get those capabilities: to build them, to acquire them, to pay for them; however you need to bring them about.
The earlier you can start those conversations, the better.