Here is something that nine out of ten conversation designers, and nine out of 10 businesses implementing chat bots, voice bots and conversational automation get entirely wrong.
I bet that the vast majority of people start all of their conversational interactions like this.
“Hi, I am [insert bot name here].”
“I can help with these three things.”
“How can I help you today?”
Or some variation of that welcome message. Introduce yourself, give the user the scope of what you're there to do and invite the user to tell you what it is that they need.
Typically, that is known as best practice in conversation design, but there's one thing that is missing, crucially missing, in that message.
That is the current context of the user.
Last night, I spoke to Anand Janefalkar, who is the CEO of Ujet. And Ujet are doing some amazing things in the customer experience space.
One of the things that Anand mentioned, which was an unbelievable insight, one of many unbelievable insights, actually, in the course of that conversation, was that artificial intelligence without intelligent automation is pointless.
AI without IA is pointless.
What that means, and this is the main thing that I think most businesses miss out on, that most conversation designers forget about, is using the data that you already have about the customer to make that interaction as personalized, as contextual and as relevant as possible.
What does that mean? Well, imagine that you're creating a chatbot, the chatbot’s going to exist on a website.
And that website is going to be plumbed into a whole bunch of back-end systems. Systems that allow you to access data, that will allow you to inform the current context of that customer. Perhaps the user is logged in, and they've recently made a transaction and they're on a returns page and they open up the bot. What is it that you think they want to talk about?
Maybe they’re on a checkout page and they've got a certain items in their basket and they open up your chat bot.What do they want to talk about there?
What if they call your call center? And you can recognize their phone number, understand what their previous interaction with the company was and whether this conversation is likely to be relevant for that.
If you use the data that you already have on your customers, either personal data, based on previous customer interactions, or contextual, based on that page that the customer is on, or the things, the activity that they've been doing previously, then you're able to provide truly personalized, truly contextual AI assistance that is definitely going to be better for customer experience.
It's easier said than done, obviously. It requires access to that data, which is one of the biggest challenges, most organizations face when it comes to any kind of automation, but the result is much, much better.
Let me know what you think.
Presented by Symbl.ai and Deepgram
Learn more about Symbl.ai: [ Ссылка ]
Learn more about Deepgram: [ Ссылка ]
Subscribe to VUX World: [ Ссылка ]
