Businesses that are using Chatbots are expanding rapidly. One of them is online shopping. In terms of traditional online shopping, customers select the product by looking at the site’s catalog. With clothing, one selects the size and color then fills out the purchase form through the chatting system. Here we introduce two cases of online shopping in America. Let’s look at the sample screen to see how the shopping is done.
Operator, an online company that sells a variety of products.
First is the case of the Operator.
Photo 1 Operator (https://www.operator.com/)
The website sells a wide range of products such as home appliances, clothing, tickets, and furniture. Figure 2 is a screen shot of a customer searching for flowers he wants to buy.
If the customer writes “I want to send a bouquet as a gift to my wife next week”, it will respond “Is there a particular flower that she likes?” If he also wrote “She likes pink flowers and I would like to choose flowers that are in season”, the Bot would select a bouquet of tulips and display a picture, price, and other details. This is a simple example, but the response will proceed automatically in the form of a Bot.
This time, here is an example of purchasing through chatting. The customer asks, “I’m looking for 3 tickets to see Bonnaro this summer” to which the Bot answers “Tickets for next Friday at 10 o’clock are available for purchase”. The Bot also asks, “Do you also need admission tickets for the camp?” at the same time. The Bot sells tickets and also makes suggestions for other items. This form of response is difficult to expect from ordinary online shopping sites. ‘Anything’ (Photo 4) is another you should take note of.
The Bot is capable of responding to a customer’s with messages such as “I want to surprise my younger sibling with a graduation gift”. The Bot will respond, “Of course. Is there anything she wants in particular?” If the customer says “She is going on a 3-day kayaking trip so it would be nice to get her something she needs for that”, the Bot will recommend waterproof speakers. This kind of response is worth taking note.
Mezi Can Show You Reviews
The next example is Mezi.
Photo 5 Mezi(https://mezi.com/)
It starts off with a request. (Photo 6)
Here the customer writes that she is looking for new sneakers and adds “I’m looking for something really hip”. Here, slang such as the word ‘hip’ is being used. The Bot responds by asking her shoe size and saying “I’ll be back in a few minutes with some fun options for you.” The sneaker that the Bot actually chose will come up along with its reviews. (Photo 7)
It can be said that the Bot well-understood the meaning of ‘hip’ as it introduced the “fashionable and hip sneakers” to the customer. (though we don’t know if they are actually sold or not) Even though the screenshot shows only one pair of sneakers, the Bot can show multiple options and let the customer choose among them. The review shown is from an actual purchaser. With the reviews, you can get product information and testimonials that you can’t get through corresponding with the Bot. It can be said that there was research done about this aspect. The final confirmation screen (Photo 8)
You can purchase after a final review. There is also information about the services after the purchase.
Changes in the Customer Response System through Chatting
Even though it is carried out by a human operator instead of a ChatBot, Daito Trust Construction in Japan is offering services for people looking for housing by using Line Talk. From consultations on renting an apartment or a mansion to tips on the moving in process, the designated operator takes full responsibility using Line Talk to advise. The image is shown below.
The operator is responding to the customer’s need by effectively using emoticons. The company is often attracting customers because they can enter what they are looking for simply and conveniently through chatting. You can say that the convenient usability of a Chat interface is being used in business just as it is. It also has another effect. Basic questions and inquiries that were not asked much in the past are seemingly increasing. In many cases, chatting allowed customers ask questions that are hard for them to ask over the phone or emails. Ultimately, Daito Trust analyzes all the factors that are contributing to the attraction of more customers and winning more contracts. Although they differ from Daito Trust’s case, upon adopting similar chatting system, some companies are achieving cost savings by automating parts of their system. Through these services, much more content to questions and trend analysis becomes possible. The ChatBot can reply directly based on the database built through the analysis of frequently asked questions and answers, and in the case where there is no corresponding database available and personal response is necessary, an operator will respond.
(Figure: The System Structure of the Operator and the ChatBot) The division of customer service between the operator and the ChatBot might make swift response and cost saving possible as well. The shopping system introduced here has been using the ChatBot for 100% of their business needs from the beginning. But it might be a good idea to use humans in the beginning, and partially and gradually move on to the ChatBot system while building experience and history during the process. Besides, it will be meaningless if 100% response by the ChatBot results in unsatisfactory customer service. Soon, it would be beneficial to consider the service content and expenses while making a plan to introduce a ChatBot. Anyhow, the Chat service has created a point of contact for new customers that was unprecedented before, allowed new possibilities of corresponding with customers and is able to meet the needs in a completely different way. We will continue to make fun ChatBot related contributions.