Adding intelligence to every customer interaction
What is a chatbot?
There’s a lot of hype these days about chatbots and how they can transform your business. Marketers at technology companies tout the value of chatbots at technology forums and business consultants sing their praises. What exactly is a chatbot?
According to Gartner, chatbots are essentially applications of voice and messaging platforms that define how people converse with digital business services and data.
A suite of technologies including artificial intelligence (AI), machine learning (ML), cloud, automatic speech recognition, natural language processing (NLP) and deep learning technologies enable chatbots to appear to behave like humans. They engage in human-like conversational experiences through voice and text that can be embedded in many types of devices including mobile phones, web browsers and chat platforms.
Why do chatbots matter to the enterprise?
The potential benefits of chatbots for enterprises—many of which are massive organizations—are significant in their ability to provide services to a range of users and groups including both customers and employees. For instance, enterprises can reduce staff workloads and increase customer satisfaction with an “always on” customer-service chatbot.
Regardless of the organization, customers always have questions. They want organizations to answer their questions quickly, care about their satisfaction and always be there for them. They do not want to be put on hold until someone is available to talk to them.
The typical AI-based customer service chatbot can scale. It is always available to greet the customer, regardless of how many are requesting help. It knows all the organization’s products and the rules about how the organization wants to communicate. Besides the ability to process multiple user requests simultaneously at tremendous speed, chatbots can help build customer loyalty and create closer relationships.
If a customer needs to talk to a human, the chatbot will hand them off to the appropriate human member of the team. Chatbots eliminate the need for customers to waste time endlessly searching through FAQs and self-serve content.
For all of these reasons and more, the AI-powered chatbot is the future of customer support.
What kind of chatbot do you need?
Once the enterprise team has completed the discovery process, it is ready to start to design and build the optimal chatbot. At this point, it is important to think carefully about the problems the chatbot will need to solve and the suite of functions the team wants to incorporate into the chatbot. Knowing exactly what the team needs the chatbot to solve makes it easier to decide the type of chatbot to build. This decision will impact both the team’s development plan as well as the end user’s experience.
Below is a description of the three main types of chatbots, their uses and variations
- Support chatbots are built to master a single domain, such as knowledge about a company. They need to be given a personality, multi-turn capability and context awareness. They should be able to walk a user through any major business process and answer a wide range of FAQ-type questions.
- Skills chatbots are typically single-turn-type chatbots that do not require a lot of contextual awareness. They are given fixed commands that are intended to make life easier. For example, “turn on my living room lights.”
- Assistant chatbots are in the middle ground between the support and skills chatbots. They need to be conversational and able to respond to just about any type of question while being entertaining at the same time. People engage with them even if they cannot perform a command because the response they provide is interesting or entertaining, or both.
What can chatbots do?
If implemented strategically, customer-support chatbots can automate many tedious business tasks, processes, and activities. However, it’s important to pre-define the roles for the chatbot. Here are six examples of specific chatbot roles:
- Improve customer service. Provide faster around-the-clock feedback for common customer questions where there is no wait time for a response and no need to search through FAQs.
- Improve response rates. Chatbots respond to 100% of messages and successfully convert visitors into buyers.
- Automate repetitive tasks. Most customers ask the same questions, so there’s no need to write or speak the same answers over and over again.
- Reduce employee workload. Enable customer support team to dedicate more time in resolving complicated customer issues.
- Eliminate clutter. Chatbots cut down on the number of emails businesses receives.
- Measure performance. Chatbots generate records of all interactions and measure customer sentiment, etc.
No matter what type of chatbots the enterprise decides to build, it is important to give them personality. People interact with the chatbots because they want to communicate in a more natural way than has been possible with previous technologies.
The handoff to a human agent
While most of the inquiries chatbots handle are resolved without a problem, there will always be a percentage of instances in which the conversation will need to be handed off to a human customer agent. For example, if a customer asks to speak to a human, the customer should be handed over immediately to a live human agent. Also, if a customer is dissatisfied with a product or service they have received they typically want to a live human agent. Chatbots should be able to recognize the different categories of customer sentiment and hand over the conversation accordingly.
By 2021, Gartner projects that over 50% of enterprises will spend more annually on chatbots and chatbot development than on traditional mobile app development. The convenience of use, minimal change management required, low-security risk and ease of deployment make chatbots one of the most favored AI technologies to emerge. Decision makers need to constantly ideate, analyze and experiment with chatbots and then strategically deploy them if they expect to improve business outcomes.
However, it takes a lot of work to turn a chatbot idea into reality. It requires a step-by-step strategy starting with goal definition, followed by choosing a channel for interaction, developing and selecting a builder platform, publishing and maintaining the chatbots.
Chatbot pilots as proof-of-concept trials can be a good starting place. Of course, the budget for the pilot will vary depending on the role (or roles) the chatbot will play and the kind of information it will need to access and communicate.
Altran’s chatbot capabilities
For any business, the goal is to create a better customer experience by providing efficient, high-quality service. Investing in chatbot technology establishes a path forward for maximizing productivity and driving the future success of the business while building stronger relationships with the customer.
As a part of our Intelligent Support offering, Altran has developed multiple chatbot solutions for a wide variety of customers. Examples include an AI chatbot deployed as a network operation center (NOC) assistant and a customer-support assistant and to automate product configuration and deployment. Altran has expertise in developing chatbots using open-source frameworks such as Rasa, as well as third-party chatbot frameworks such as Amazon Lex, Microsoft Bot Framework and IBM Watson. Altran recently made a strategic investment in an internal Innovation group that includes a talented team of engineers who are developing virtual-assistant solutions and contributing to the open-source community.
Speak to a Human about AI Chatbots.