When was the last time you had a good experience with chatbots? Thought so.
It is, therefore, a little alarming to read Juniper Research’s prediction that AI based chatbots are set to have 22 billion interactions in retail alone, by 2023, just four years away.
It is entirely possible that the cost savings of implementing chatbot technology will be significant – $439 million in annual savings by 2023, according to Juniper.
It is also possible that chatbots will also increase sales, although these will probably not be ‘new’ sales but cannibalised from other channels.
One of the few things we can agree upon is that excellent customer service is everyone’s priority, no matter what arena you are in. And right now, chatbots are harming the customer experience not enhancing it.
The technology will have to improve dramatically in a very short time frame in order for this to change significantly. There is talk of improvements in Natural Language Processing but it is still a long way from perfect.
And perfect (or very, very close to perfect) is mandatory. When Optical Character Recognition (OCR) was ‘the big thing’ back in the late 1980s and 1990s companies would boast that their system was now 90% accurate. But in the world of OCR 90% accurate means that 10 in every 100 words was inaccurate and that was a joke if you were dealing with huge documents.
And OCR is only about words, chatbots need to be about ‘human’ conversations (have you ever listened to a recording of a conversation, it is nothing like we believe it to be) and answering questions in a friendly and efficient way. And understanding 100% of words. There is the story of someone asking Alexa to ‘add gym’ to the list of things to do. Alexa responded by saying that it had ‘added gin to the list’.
It obviously makes sense to integrate technology into processes but you have to wonder about the financial pressure on companies to automate anything that isn’t bolted to the floor and ending up, as happens in tech industries, that we overcompensate, over digitise and end up achieving the opposite of what we wanted to achieve.