It was interesting to be able to experience an insider side of Chatbot at the previous class by creating a Chatbot using Dexter. It is true that so many questions can be answered by Chatbots which enables a company to reduce time and cost of customer service representatives by automation of process. For example, Ele-me, China’s largest food delivery company receives two hundred thousand support requests between 11–2pm on weekdays. 90% are ‘My food hasn’t arrived yet’ and ‘My food fell out of the container’. Chatbots can function as a customer service rep. I could create a message like “I am sorry to hear your food hasn’t been delivered yet. I am sure you are hungry! To show you an apology I’d like to offer you 50% off any entrée on your next purchase.”
Chatbots are useful in a range of industries, from simple banking transactions, to eRetail customer service inquiries.
Juniper’s new research, Chatbots: Retail, eCommerce, Banking & Healthcare 2017-2022, forecasts that chatbots will be responsible for cost savings of over $8 billion per annum by 2022, up from $20 million this year.
A data shows that people value messaging which proves Chatbots have a promising future.
1. 51 percent of people say a business needs to be available 24/7
People want instant availability 24/7 without waiting for a store to open and going there.
2. 45.8 percent would rather contact a business through messaging than email
People want to move away from email with “info@…” and wait for 24-48 hours to get a response. They rather have an instant yet personal conversation with messaging.
3. 49.4 percent would rather contact a business through messaging than phone
I find it very hard and stressful to reach a real person without going through several options first on the phone. Then I have to hold for a long time to talk to a rep. Of course I easily miss an open window. I have to reach them between 10am-5pm M-F for example, which is not easy for me because I work during this time.
source: The future of Chatbots
Many simple tasks can be handled by Chatbots. However it has to be natural like talking to a real human that can deliver a seamless user experience. So how do developers create Chatbots that are more natural and flexible?
In combination with the power of NLP, developers use functional languages such as .NET or Java to create decisions-trees or slot-based algorithms that lead the user through a predefined conversation path.
They now use “declarative programming.” to consider natural language that can deal with a situation such as when people change their mind in the middle of a conversation. This solution is 16 times faster and twice as cheap.
There is still a room for improvement. However the technology is getting better and if we can implement it successfully, users will become closer to your products than ever before, benefiting from being better understood on their own terms.
Chatbots: Key Statistics