Conversational AI is a scary topic and a scary technology to introduce into your business. Getting it right first time is paramount when introducing a technology to a business. The same applies for a new vendor introducing a new element to their portfolio. Yet we seem to be nearing the point where is it readily accessible and a standard part of a contact centre offering.
What’s driving the industry?
Productivity. The need to get things done in a timely manner. Enterprises realise there is a lot to be gained by introducing bots and machines to do the mundane tasks that humans spend their time on.
When looking at chatbots standalone, we typically run into brands like Live Person, LiveChat, Amazon Lex, Dialogflow, IBM Watson and Bold360. With different backgrounds – live chat, artificial intelligence and voice recognition, these players all have their own unique offering. Live chat is evolving and there’s no doubt these vendors have a big say in that.
Contact centre AI
When it comes to contact centre, there are some vendors leading the way. By bolting on, integrating or even buying conversation AI software, there were a few vendors that immediately came into my head when I was asked to write an article on conversational AI.
RingCentral
When RingCentral bought Dimelo last year, they added a customer engagement play to their contact centre catalogue. Recognising that by 2021, 15% of customer service engagements will be handled by AI, this seems a sensible move.
Dimelo, now operating as RingCentral Engage, provides conversational AI as the first line of the battle for customer engagement and cite their bot can provide the following for small businesses up to enterprise level:
- Gather and qualify leads that come to your website by answering their questions and directing them along the sales funnel
- Collect useful data about your customers from their conversations, which you can use to alter and improve your service and sales strategies
- Keep customers engaged by providing them with plenty of reliable ways to contact your brand and answering any questions they might have about your product or service
- Free your employees up to complete more important tasks by taking over and automating repetitive and monotonous jobs in the contact centre
Dialpad
Dialpad is another company that has been on the acquisition trail. They picked up TalkIQ last year which came with a handy voice AI offering.
Customer service reps can benefit from prompts and reminders when chatting in real time with an agent. For example, if the customer has expressed interest in buying an apple as well as a pear, but the agent has only sold a pear, the voice AI will prompt the apple sale too. This benefits both parties. The business gets both sales and the customer doesn’t have to call back and place another order.
As this is happening, patterns and trends in conversations are automatically being analysed to improve pitches and scripts. Rather than losing the caller, the problem can be resolved there and then.
https://youtu.be/dEkA84tVX5c
Cirrus Response
I have previously sat in on a contact centre demo where Kris Graham, Business Development Manager at Cirrus Response, blew the room away. Readily situated on the customer’s website was their conversational AI chatbot. It could recognise what the demo customer was asking and respond rapidly. It even searched the website for information that was only available there. The mind blowing element was that Kris hadn’t done any back-end configuration with the customer and the bot was getting the answers right.
As I searched for further conversational AI players, the first sponsored post on Google was for Ami. Naturally, I began a web chat with the conversational AI. I asked this Ami if she was the same Ami that Cirrus Response sell, remembering that was the name of their AI bot. Remembering what a great experience I’d had with previous Ami, I was expecting a rapid response with the information I had requested. Here’s Ami’s reply:
“Did you know 80% of questions that customer service agents get asked can be answered by information on their websites? I learn by reading pages on my client’s websites, and unfortunately, the question you asked is not available here. Let me take some details and one of the team will get back to you with an answer!”
Safe to say I wasn’t as impressed as I was when I saw the original demo with Kris last year.
Live conversation data
Louis Halpern, Chairman of Active OMG, the company behind Ami, called me swiftly after and explained that Ami used to be white labelled by Cirrus Response, hence my confusion. He then explained why Ami couldn’t answer my simple query and told me a little about how Ami works.
“Ami uses live conversational data to solve a complex query like finding the price of your flight or finding the postcode of your nearest shop location. So, if the information is on the website, it can find it, make sense of it and relate it back to your question”
Louis also told me about Ami’s reading function. “Ami reads the website and makes assumptions. For example, Ami understands the word shirt and knows the website sells shirts. In the background, the AI software is making sure these assumptions are correct so Ami can make the best shirt recommendation based on the conversation she is having”.
My faith restored in Ami, Louis told me about a further addition to Ami’s armour. Once Ami knows you want to buy something, she is looking for the outcome. If, after saying you want to buy a product, then digress and start talking about something else, Ami will bring the conversation back to the original product request.
Conversational AI uses cases
The use cases for conversational AI are clear. For queries that include information accessible on your website, conversational AI can remove the human resource to stand by and power your live chat.
Customer service queries and early sales signs can be reacted to in an instant. And with chatbots, there will never be a queue – so your customer experience is already benefiting.
Productivity is easily gained through implementing conversational AI. The technology is ready as long as you understand its limits. But will YOUR customers be so understanding?
