零售業新革命!Chatbot如何創造零售業新價值? A retail revolution: How chatbots create new value in the retail industry

越來越多零售業者導入Chatbot(智能客服)分擔客服人員的負擔,世界上最成功的統計資料庫之一Statista也公布了哪些產業使用Chatbot最能讓消費者接受,當中零售業獨佔鰲頭,成為唯一超過三成的產業。
More and more retailers are introducing smart customer service chatbots to ease the burden placed on customer service staff. Statista, one of the biggest statistical databases in the world, recently announced the industries using chatbots with the highest acceptance rates among customers. Out of all the industries, retail takes the lead, and becomes the only industry with an acceptance rate over 30%.

而根據美國行銷顧問公司Invesp的創辦人在官網上公布的一項數據:在過去一年內,全世界有超過67%的顧客透過Chatbot滿足了他們在商品服務上的需求。此數據體現出了透過Chatbot來取得服務已經是大多數人的習慣了,但今天當零售業者在導入Chatbot時,應該要如何建置,才能在減輕客服人員負擔的同時,還能讓Chatbot發揮出其他價值呢?


行銷 V.S. 服務,服務型Chatbot創造更好消費者體驗
首先先來搞清楚,在零售業者導入Chatbot,可以先區分清楚自己想做的究竟是專為行銷專案做的Chatbot,還是作為智能客服的服務型Chatbot。
一般行銷型Chatbot由於主要是配合行銷專案所產出的,大多數是配合貼文主題或是抽獎活動,並以關鍵字抓取或是選擇性牌卡的方式去建置的,所以在一檔行銷活動結束後,Chatbot價值的週期便會結束,而不能持續提供產品相關的服務給消費者。
所以,相對於行銷型的Chatbot來說,服務型的Chatbot可以發揮的價值卻不僅於此。以零售業來說,大多數的服務型的Chatbot,像是許多公司會有的智能客服,除了可以用來讓顧客做一般的資訊查詢外,若能夠透過API或是其他方式,將Chatbot串接到公司內部的系統的話,顧客也可以利用Chatbot去「完成服務」,而並非只是制式化的查詢資訊。

你的Chatbot夠聰明嗎?用NLU打造聰明大腦
Chatbot除了可以協助客服人員處理基本需求,也可以幫助企業創作另外的價值。然而,Chatbot既然有這麼多優點,為何在仿間還是會有很多人聽到很多人說:「智能客服一點都不智能!」、「它根本就沒聽懂我在說什麼!」
以往的Chatbot大多數是透過關鍵字比對來建立,因此消費者若使用了非預設問法的問題詢問Chatbot,那Chatbot就會聽不懂,那除了關鍵字比對外,還有什麼方法可以讓Chatbot變聰明呢?
以SysTalk.Chat為例,SysTalk.Chat使用NLU(Natural Language Understanding, 自然語言理解),進行意圖分類(Intent)、實體識別(Entity),讓Chatbot理解對話內容,正確掌握消費者服務需求與資訊,以供後續進行對應的服務流程。
以零售業者來說,當消費者想要詢問該零售業者的營業時間時,若Chatbot使用NLU去訓練,無論是依照流程詢問「我想要查詢營業時間」,或是消費者詢問Chatbot「十點會開嗎?」,Chatbot都可以理解消費者是想要查詢「營業時間」,因為這兩句話背後的意圖是一樣的。但若使用關鍵字比對,可能一定得輸入「我想要查詢營業時間」,消費者才能得到想要的答案,但這樣的問法卻一點都不符合消費者一般的詢問習慣,十分不人性化。
創造更能理解消費者心思的新零售,完成更多服務貼近人心的服務!一如Statista的調查:64%的企業都相信Chatbot能夠為顧客完成更多的服務,對於零售業來說,Chatbot不改只被當作行銷、更是要深入消費者心思,提升更好的服務品質。

According to data published on the official website of the American marketing consulting firm Invesp, in the past year, more than 67% of customers around the world have met their needs for products and services through chatbots. This number indicates that people have already developed a habit of using chatbots to acquire services. But in the current retail landscape, how should chatbots be built and implemented in order to reduce the burden placed on customer service staff while allowing chatbots to generate new value?

A better consumer experience with service chatbots
To implement a chatbot in a retail business, you must first decide whether you want a chatbot specifically for marketing support or a service-based chatbot for intelligent customer service. Marketing chatbots are mostly implemented in conjunction with marketing projects and built in accordance with the theme of the social media post or promotional activity planned. Once the marketing activities are over, the lifecycle of the chatbot value ends, and product-related services cease to be provided to customers.
In comparison to the marketing type, service chatbots create a lot more value for organizations. In the retail industry, most service-oriented chatbots, such as those employed for smart customer service in many companies, can enable general information inquiries for customers via APIs or other similar connection methods. Beyond standardizing information, customers can also use chatbots to “complete” their services.

Is your chatbot smart enough? Build a smart brain with NLU!
Beyond assisting customer service staff in handling basic tasks, chatbots come with many other advantages and can also generate additional value for companies. Why, then, do we often hear people claim that “smart chatbots are not smart at all” or “they do not understand what I am saying”?
In the past, most chatbots were established with keyword matching technology. If consumers were to ask the chatbot questions that are not preset, the chatbot would not understand them. In addition to keyword matching, there are other methods which can be used to make chatbots smarter.
Take SysTalk.Chat as an example. SysTalk.Chat is an AI conversational platform which uses natural language understanding (NLU) to perform intent classification and entity recognition. This allows the chatbot to understand conversation content and correctly grasp customer service needs and information, thus identifying subsequent corresponding service processes.
For a detailed example, think of a customer inquiring about the business hours of a retail store. If the chatbot is trained with NLU, whether the customer asks “Will your store be open at 10am?” or “I want to know your business hours,” the chatbot will understand that the customer wants to know the “business hours," because the intention behind these two sentences is the same. On the other hand, if the chatbot is trained with keyword matching, the customer may have to say “I want to know your business hours" exactly to get the answer. This kind of input method, however, does not suit the general inquiry habits of customers and lacks the human touch of a smarter chatbot.
Create new retail value with a chatbot that can better understand customer minds and complete more services successfully. As the Statista survey indicates, 64% of companies believe that chatbots can provide a wider variety of services for customers. For the retail industry, chatbots are not only used as marketing tools – they must also be deeply rooted in consumer behavior to improve service quality.

發表迴響

在下方填入你的資料或按右方圖示以社群網站登入:

WordPress.com 標誌

您的留言將使用 WordPress.com 帳號。 登出 /  變更 )

Google photo

您的留言將使用 Google 帳號。 登出 /  變更 )

Twitter picture

您的留言將使用 Twitter 帳號。 登出 /  變更 )

Facebook照片

您的留言將使用 Facebook 帳號。 登出 /  變更 )

連結到 %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.