I often wish that my company did something easier to explain. We make bricks, we grow plants, we brew beer. Nope we, wait for it…. we…. here it comes... we do... Proactive Conversational AI.
Snappy huh?
Now if you used your old metaphorical gran as the arbiter of what constitutes a good answer, she would have given me an epic fail. So, Iet’s try that again. ‘So Nan, ContactEngine starts a conversation with our client’s customer (she’s not raised an eyebrow yet and nodding approval….), invites them to reply in a very human like way (she’s looking a bit puzzled but still all there) and when we get their reply we answer back in a way that they understand’.
‘Oh’ she says, ‘you just talk to people’?
‘Yup, that’s it, we talk, we listen and we reply and we do that using computers, not people’ I say. Satisfied, I make the metaphorical tea and butter the metaphorical scones.
Now the problem here is that loads of companies have now taken to talking about Conversational AI and why not? The advances in Natural Language Understanding have been extraordinary in recent times and we get closer and often exceed Turing’s AI test – whereby it’s often not possible to detect whether it is a human or a machine replying.
So what’s my gripe?
Well the thing is nearly everyone ignores the P Word – Proactive. That is the difference between good service and poor service, the difference between frictionless and friction inducing engagements. Simply put, that is where Conversational AI is destined to go.
You see, we have a problem. Our details as a customer are held in a variety of different systems inside the organisations we transact with – it might be inside a billing system, a Customer Relationship Management (CRM) system, it might pop up inside a workforce management system, etc. etc., and there is one unifying aspect of these systems – they are sold and bought on a ‘per seat’ licence basis.
‘What’s that you said, son? A per seat what?’ asks Nan.
‘Well that means all the people a company employ that talk to a customer need software to present the customers details and the software companies that provide it charge based on the number of people that use it’ I explain….
‘But son, if what you do doesn’t need so many people – doesn’t that mean that these companies that sell ‘per seat’ are knackered?’
‘Eh yes I suppose it does Nan’ is my reply…….
And that is the nub of the problem, if a company has a vested interest in keeping people in front of their software, then they are not likely to embrace a service that eats away at the amount of people needing to do that work.
By proactively engaging customers in conversation, the software market needs to change in a very similar way to the revolution in hosting. Pre ‘cloud’, if you wanted to host a website you needed to literally buy the servers on which to host it. Worse, if you didn’t want to share that server, you had to buy your very own box - maybe several. All rather expensive. One of my past businesses did just that and spent around £500k just to host some relatively simple online conference software. Then, along came an online book seller called Amazon who had loads of spare capacity when it wasn’t selling paperbacks... and out popped The Elastic Compute Cloud (EC2) from AWS in 2006. Bingo! You just paid for what you used. The more successful you got, the more you paid. Nice.
That is where Conversational AI will go. Just think about it, if you are picking up the phone to a call centre, it’s mostly because stuff has gone wrong – a delivery is late or broken, you have a holiday to cancel, your washing machine is on the blink, your insurance premium is too much, you have a routine dental appointment to make or an eye test to sort.
Proactive conversations stop that irritating task – they let you know that your appointment is due, a pandemic might just have changed your travel plans, and why your premium has gone up.
Trouble is – if your business model is so pre-2006 and you just want to sell ‘per-seat’, why would you want Proactive Conversational AI? Well it seems many don’t – so the investments being made in AI are all about diverting inbound calls and requests, often from already irritated customers, to a poorer ‘chat bot’ experience so that customers can ‘self-serve’. ContactEngine on the other hand reaches the customer first, starts the conversation and invites a human like conversation – that makes the customer happy, makes the call centre agent happy (as they get to deal with more complex issues) and saves companies lots of money.
So as my old gran would have said, ‘looks like a communication revolution is underway, go change the world son, one pro-active conversation at a time, and pass me a scone would you’?
