I got new Amazon Alexa as a Christmas gift this year.
To handle our family calendar, we have created one shared Google account. Initially I thought that I could use it with a new member. The reason behind this approach was that I didn’t want to connect my own Amazon account to this service as all the members of our family (and also, possibly, visitors) would use it. Quite early on, it became evident that to get all possible integrations working, it would be much easier to have same email in as many integrated services as possible.
For this to happen I had to re-register my phone number to another Amazon account. This turned out to be much more difficult than one would have thought. I couldn’t just register the number with a new account as it was registered for the other account. I searched through the interfaces for how to deactivate the number from the old account. Eventually I ended up just removing it; but that didn’t help either. Eventually, I found a place in the user interface where I could request help.
Based on my request, I got an answer. Amazon requested that I call them or have a chat session just to verify that it is me who owns the phone number that I’m updating. I think that this is good security practice.
“As they have developed a good Artificial Intelligence as part of Amazon, I decided to try the chat option.”
The first chat agent requested more information about my case and quite quickly forwarded me to another chat agent. This same pattern repeated a few times before one agent requested me to verify that I have the phone number by sending me a code in a text message. Eventually I was passed on to the next agent six or seven times until I eventually arrived at the agent who could solve my problem. After they had fixed my problem I began to analyze the service.
Taking a new look at chat bot service development
What are the key key things that we learned from this story in order to understand future chat bot service development? If there was an AI robot behind of this service, it wasn’t very good. They pestered me several times to input data that I had already supplied. As a user I’m obviously more interested in the result than the process of how we got there.
Even if these ‘assistants’ were not chat robots, the agents were not very good. It’s these large technology companies who would have the capability to create these chat bots with artificial intelligence that gives a better customer experience saving both my and the agents’ time.
Customers know the service has their information and they expect services that have been created to use this data. Given that I had already defined the problem by clicking the help link on a certain page; I would have expected them to do just this.
Based on service development, the following considerations should be taken into account:
- Creation of a customer journey map to better understand the customer.
- Collection of real time data for the customer journey.
- Collection of real time customer feedback.
- Creation of automates solutions to typical problems.
Do you have more ideas on chat bot development? Add them to the comments below.