More and more companies have a digital employee a.k.a. chatbot. Most are active on websites, but bots can also be used in apps, messaging, on social media channels, and in WhatsApp. A bot is an attractive way for many sectors to quickly answer simple questions without the need for an employee. Everything you find on an FAQ page can be programmed in a bot.
Such chatbots often work with pre-programmed conversations. The customer asks a question, and the bot finds the right answer. When in doubt, the bot shows various options so the customer can choose which answer best suits his or her question. If the bot can’t find what the customer is after, the conversation can be passed on to a real employee in a natural way. Without the customer having to repeat his question.
Nevertheless, a recent panel survey* by the Dutch Consumentenbond (Consumers’ Association) among 10,000 panel members shows that 66% of the respondents are not satisfied with bots. Why? Because they often received no or only a partial answer to their question. And it was difficult to switch from the bot to a real agent. Conversations with bots were also experienced as impersonal.
Bots are often intended for simple questions, but that does not mean customers ask only those. It’s important to manage expectations and clearly explain what customers can use the bot for. And continuous improvement is also necessary. It’s not a matter of programming, launching and that’s it. The learning curve starts after activation. The more advanced bots can partly do this themselves using Artificial Intelligence (AI) and Natural Language Processing (NLP). But on top of that, it’s just as important to have a team that regularly analyzes conversations and add new questions and answers. This helps bots determine what customers are after and provide the right assistance.
There are many bots that have trouble with transferring a conversation to an agent if the bot can’t find the right answer. It sounds like an easy task but that’s not always the case. Most of the times this happens because the bot is a separate tool and not part of an omnichannel platform. To transfer a conversation to an agent, the bot needs to be connected to the chat tool that the agents use. And in case of switching to a phone call for example, an integration with the telephony infrastructure is required. Ideally the bots should also have access to the CRM software so they can gather all required information for the agent, and he/she can easily pick up the conversation. When choosing a bot, these kinds of integrations are important to keep an eye on.
You can also use chatbots to request necessary data prior to a conversation with an agent. This shortens the waiting time. Based on that data, the bot looks up the customer data in the CRM software and the agent can respond more quickly to the real question. According to Gartner**, this can reduce call handling time by 25%. This potentially yields significant savings on an annual basis.
Cost savings are, in addition to increasing accessibility, one of the most important reasons for using chatbots. Your agents are the largest cost item in customer service, so the more you can automate smartly, the more efficiently you can deploy those employees. Having a bot handle frequently asked questions also makes the work of the agents more rewarding. The simple repetitive questions and tasks go to the bots, giving them more time and focus for complex and more challenging questions.
Gartner expects 40% of customer queries will be handled by self-service or a bot by 2025. For this to happen, it’s necessary that bots are part of the omnichannel mix and have integrations with CRM software. Of course, bots still have a lot to learn, like recognizing emotions. But research by the Dutch Consumers’ Association reveals that most of the companies they interviewed already see that many questions are handled by the bots. So yes, if properly implemented and maintained you can already benefit from using high-quality bots. And that can only increase with continuous improvement.
Want to find out if a chatbot can work for your customer service?
Get in touch. We’ll be happy to take you through various options and help you determine which best fits your needs.
* Research Consumers’ Association magazine April 2021
** Gartner Global Market Trends: Contact Center Innovations Promise Better Customer Experience, October 2020
This article was originally published in Dutch on Emerce.nl