Chatbots serve your customers better
In these customer centric times what better than a way to improve your customer service. In these P&L heydays what better than a way to reduce your cost. A chatbot is a great way to do both. Increasing customer satisfaction while at the same time reducing the cost of serving your customers. Please note this applies as much to ‘real’ external customers as to internal customers. Too good to be true? Let’s examine the technological developments, some real life experiences, perspectives on how to implement chatbots, and their bright future potential.
From zero to hero
People above 30 years old will remember Clippy the office assistant Microsoft included in Office for Windows from 1997 till 2003. Any user would try to silence Clippy as soon as possible as it was of absolutely no value. It would never come up with an answer for the problems one was experiencing.
The days of Clippy are over and since a few years chatbots are becoming a very useful, always available and highly scalable tool to support users in getting what they need quickly and instantaneously.
The biggest progress comes from AI and its ability to understand natural language. The user no longer has to describe his or her problem in exactly the same wording as the tool makers were expecting. AI is able to interpret the natural language pretty accurately and identify the intent of the user. Once the question is properly classified (“where can I find the print button….” e.g.) the chatbot feeds back the right answer from a knowledge base or executes an action (“buy this product…”).
The world’s organizations are discovering the benefits of chatbots fast. Research&Markets.com reports ‘the chatbot market size is projected to grow from USD 2.6 billion in 2019 to USD 9.4 billion by 2024, at a CAGR of 29.7% during the forecast period’. Gartner says Artificial Intelligence (AI) will be a mainstream customer experience investment in the next couple of years. 47% of organizations will use chatbots for customer care and 40% will deploy virtual assistants.
Implementing chatbots
There is a lot of magic talk about self learning chatbots. Don’t be fooled, with today’s technology creating and improving a chatbot still requires a significant amount of manual work:
- Defining how and where to use the chatbot in the customer journey
- Defining the intents correctly
- Designing and implementing the tone of voice of the chatbot
- Creating or collecting the knowledge articles
- Keeping the knowledge articles up-to-date
- Integration the chatbot with backend systems to execute tasks (e.g. “reset my password”)
- Analyzing interactions and improving the answers of the chatbot
Although getting the chatbot conversations right is definitely possible today, it is not a given. Forrester predicts that even in 2020 four out of five chatbot-based customer interactions will continue to flunk the Turing Test. Spiceworks provides an overview of the most common errors.
A solid implementation methodology, thorough testing, reasonable expectations, picking the right conversation cases, and an experienced partner are amongst the key success factors for implementing meaningful chatbot conversations.
Advancing technology will make things easier. For example, improving chatbot conversations is increasingly supported by strong analytical tools and automatic improvement suggestions. Also, technology is now enabling the automatic creation of new intents by scraping websites or uploading (somewhat structured) pdfs. This is e.g. part of Google Cloud’s Contact Center AI product. More and more chatbot vendors are including predefined workflows and connectors to make it easier to integrate with backend systems for the execution of identified tasks. These kinds of features reduce chatbot implementation and maintenance effort.
The next frontier, Voice is becoming mainstream
Biggest advancement however is today’s fast increasing maturity of voice technology. Where traditional chatbots require users to type to interact with them, the new wave of chatbots (or perhaps better, Voicebots) are able to interact by voice. Most mobile phones already have an assistant like Siri or Google Assistant installed, and more and more households have Alexa or Google Home (25% of US households in 2019). While American English still performs best it is quickly followed by mainstream European and Chinese languages.
Humans are very comfortable and skilled in using voice to interact with each other. Now computers can do this too. Today many organizations have implemented e-commerce, webcare and chatbots to make customer interaction easier but still operate large call centers with agents talking to customers. This is expensive for organizations and often little appreciated by customers (wait time and frequent hand-overs). Voice technology promises to speed things up for customers as they can be served immediately anytime of day no matter the call volumes, to lower costs for organizations and to augment quality of work for agents as voicebots are likely to deal with routine tasks first. So get it right, and the future is bright!