5 Rules of Chatbot Success According to Slevomat
Slávek, Slevomat’s chatbot:
Up to 23% conversion rate during peak season & 19,000 chat conversations monthly
The Chatbot Era
In recent years, the development of chatbots sky-rocketed. Not too long ago, the massively expanding chatbots were criticized for their ineffectiveness or impersonality. Nowadays, chatbots build on the powerful technologies of artificial intelligence (AI), machine learning (ML), natural language processing (NLP), or sentiment analysis, which gives them the ability to respond effectively and naturally, similarly to a human conversation.
In 2021, the use of chatbots has spread to areas of insurance, banking, healthcare, e-commerce, and retail. This year, chatbots are expected to be introduced into offices, so that they can improve the work performance and experience of employees. In 2022, Gartner predicts that 70% of white-collar workers will communicate with a conversational robot to speed up their work.
Thanks to chatbots, companies can scale their customer services without increasing operating costs. With the rise of cutting-edge technologies, even customers are more open to communicating with a chatbot, which offers them the opportunity to resolve their questions and problems immediately. The importance of chatbots is supported by statistics claiming a high rate of chatbot preference over a conventional email or a phone call. *According to Adweek, 65% of users prefer a chatbot or other automation technology to an email or a phone call for answering their questions.
Slávek came, saw, and conquered. On the second attempt, but absolutely.
Slevomat is among the largest Czech and Slovak travel and experience retailers, with 300 employees. In the Czech Republic, it has 4 million registered customers, to whom Slevomat offers exclusive offers in collaboration with freelancers or large multinational companies. Every day, it helps more than 1000 customers, strategically emphasizing perfect customer service, with the help of the Spinoco application.
The very first Slevomat chatbot was introduced in 2019 with the vision of improving the speed of handling customer requests. It was similar to an IVR tree – a simple decision tree on the Messenger platform, without the use of artificial intelligence. For customers, it was not much of a success – they usually opted for a human operator even when the offered options included their specific topics. The period of the first chatbot’s use was cut short by the pandemic when all customer communication was handled on different communication channels.
In the summer of 2020, Slávek returned back to the game – significantly innovated with the goal of high effectiveness, solution reliability, customizable design, internal chatbot management, and integration with internal systems. In just a few days, Spinoco provided Slevomat with their own chat plugin and an integrated solution using advanced artificial intelligence (IBM Watson), ready for content production.
Nowadays, Slávek can recognize up to 120 search queries and more than 130 dialogs. It can help the customers with using their purchased products and vouchers, or it can help with the right choice of a new service or product. It offers relevant trips, experiences, procedures, goods, or gifts.
The success of Slávek is supported by data –
by the end of 2021, it handled up to 19 000 chat conversations per month while reaching up to a 23% conversion rate during peak season. From the former advice tool, it made its way up to a reliable sales tool. Its success rate ranges from 90% to 93% of automatically handled requests.
No pressure, no diamonds
It is not only the advanced artificial intelligence that is behind Slávek’s success - a patient and meticulous trainer is needed as well. Slevomat dedicated a specific colleague for chatbot management - and couldn’t have done better. The chatbot solution does not require significant technical skills or technological knowledge, yet it is a time-consuming process, especially during the launching period of the chatbot.
In Slávek’s first months, it took 3 to 4 hours a day for the chatbot trainer to check the conversations, edit and add keywords, or complete dialogs. The same amount of work is needed when doing larger changes in offers or communication, or when entire new topics and dialogs are being created.
How To According to Slevomat
Are you considering introducing a chatbot in your company? Here are a few tips and tricks you might need along the way:
1.If you don’t train it, it won’t learn
Artificial intelligence has to be trained – to recognize keywords, expressions, phrases, and complex questions – and to respond accordingly
2.Let your customers take the lead
When creating the content, you will need a solid base covering the essence of your business – but you do not need to cover absolutely everything right at the start. Let the customer be the leader – watch what and how they ask (whether it be in the chatbot directly, or via Google Analytics), and further edit the chatbot nodes based on the observations.
3.Forget about cutting the human operator off
To a certain extent, there are always going to be complex tasks and questions that require to be solved more reliably by a human colleague. In the same way, there are always going to be customers who prefer talking to an operator.
Do not train the chatbot by yourself – use your internal systems. If you use the right customer information that you already have, the chatbot’s performance will increase significantly.
5.Do not expect perfection
If you don’t plan on using the chatbot only for limited and routine tasks without major changes over time, do not expect its performance to be perfect, especially not in the launching period. In this case, the journey is the destination. Moreover, it is determined by the customer who is the leader in the communication. This is your biggest challenge, but also the biggest advantage.
We will be happy to help you.
Do you want your chatbot to be your reliable help?
We will plan the best solution for you, considering your needs and options.
Get in touch, we look forward to working with you.