This blog has little bit of everything...

Friday, February 10, 2017

Why Will Chat BOTs Fail To Replace BPO Agents?

You can't ask a fish to climb a tree or ask a pig to bark. But this is exactly what most IT giants are about to do. Let us understand how...


Today as a customer when you need 'service' you call the customer service desk or ping the customer service agent on instant messenger. A portion of the phone/ call support agent's work has already been replaced by the IVR machines but the instant messaging service-agents are not replaced by the chat BOTs yet.

Leading research and IT companies have forecasted that the next 2 years will be the years of digital transformation and in this period chat BOTs will replace humans in almost all forms of first level instant messaging interactions.

But is this for real? Will we be able to create intelligent, self-learning chat bots that will be able to succeed with the help of natural language processing (NLP) to a level where they can replace humans?

The answer is Yes and No. Yes because a miniscule amount of work will certainly be taken over by the BOTs and No because there is a limitation on how far the BOTs can go. Getting a BOT to respond to customers/ users may work at a very basic level, for example, the BOT may perform tasks like...

  • Gathering information from the customer/ user
  • Raising tickets or cases on-behalf of the user
  • Sharing URLs or information or status based on the keywords entered by the user. 

Chatbots in banking, chatbots in healthcare are few examples where chatbots integration with messengers(IM) has become reality. However, the above scenarios may be applicable to only a limited set of use cases. At best it may serve as an equivalent to the IVR system for receiving calls, that is, the BOT will route the customer to the right set of information or raise a case/ ticket as mentioned above. It may not work as soon as the customer/ user states a problem which requires analysis or in scenarios where 'human' behavioral skills are required to understand and manage the customer issues/ expectations.

For example, BOTs will not be able to...
- Take ownership
- Be Proactive
- Identify when a certain issue may be on the verge of escalation
- Know Implication
- Follow-up / Identify and remove road-blocks
- Ask for help

There is one more aspect that seems ignored at this point, that aspect is of 'user adoption' and 'user acceptance'. This requires real attention. No one knows how the customers/ users will respond to a situation when they get to know that they are not chatting with a human but with a BOT which is trained to support only few scenarios.

Having said the above my fear is that if not used wisely BOTs may in turn give rise to...
- Confused Users/ Customers
- Frustrated customers/ Reduced CSAT
- Increased cycle times of issue resolution
- Potential loss of revenue

The reason why I say so is because we are only looking at one side of the story. We are talking about humans being replaced by BOTs on the 'response' side of the chat but we are not at all looking at the humans on the 'request' side of the chat window. By 'request' side I mean the users/ customers/ or people who are facing problems. For the user and customer groups it will be a huge culture change. The success of chat BOTs will depend largely on how the culture change is managed. Hence, while driving operational excellence through digitization the IT giants should not ignore this 'human' side of the reality. If they do, it may backfire at them. Giving the command of frontline support to non-humans is like flirting with danger. Beware!

I hope you find this information helpful. If you are more than 18 years old you would also like to read this post How Much Money Would One Save In USA?