Technological advances such as the Internet and mobile connectivity have drastically reduced the cost that is associated with information retrieval - allowing customers to wander through various company websites, look at different vendors, and compare a wide range of products with each other with the click of a button.

This reduction in switching costs has pushed businesses to implement marketing strategies which are heavily focused on capturing and keeping customer attention as much as possible. One of the key aspects of such a strategy is the deployment of an intelligent conversational chatbot which is able to provide website/application visitors with fine-grained information, custom-tailored nudges, and top-quality after-purchase service.

Such human-chatbot interaction is rapidly growing and is becoming one of the most important types of interaction between customers and businesses. Even more, global research and advisory firm Gartner estimates that more than 85% of customer interactions will be handled by artificial conversational chatbots by the end of 2021 - stressing the demand for sophisticated conversational chatbots.

But where are we now? Are chatbots nowadays capable of providing the much-needed conversational depth that customers are looking for - or are we forever stuck with being redirected to customer support for our truly complicated questions?Conversational chatbots utilize the power of Artificial Intelligence and Natural Language Processing (NLP) to mimic human conversation. However, depending on the level of sophistication of the underlying technology - we often differentiate between different types of chatbots.

The first type - often referred to as Conditional or Simple chatbots - have been programmed in such a way that they are capable of detecting certain keywords or text fragments and, based on these keywords, provide their conversational partners with pre-programmed answers. Whereas these chatbots are rather simplistic, they are heavily used in smaller companies to accommodate the first customer interaction and provide answers to frequently asked questions.

The second type - often referred to Advanced chatbots - are able to answer customer questions based on conversational data. Instead of searching for pre-programmed keywords, advanced chatbots use Artificial Intelligence to analyze incoming text fragments in order to more accurately determine a person’s needs and formulate better responses. However, even the AI-based Advanced chatbots are still a long way from being able to impersonate a skilled help desk employee - making the technology insufficient to capture and keep the much fought for customer attention.

The main reason for their shortcomings is the fact that the chatbots discussed above are not able to capture and recognise the context in which a conversation is taking place - making it difficult to provide customers with granular and thoughtful responses. However, this is exactly what the new kid on the block - Contextual chatbots, often referred to as Virtual Agents or Smart Chatbots - is trying to solve by taking into consideration a conversation's context.

As discussed in the previous section, unlike other types of conversational chatbots, a contextual chatbot is capable of capturing and recognising the context in which a conversation is taking place. But what is meant by the context when talking about online conversations taking place between humans and virtual chatbots?

First of all, Contextual Chatbots are linked to the website’s or application’s database. This allows them to retrieve crucial information about the person they are conversing with - including the person’s name, address, or previous purchases he or she has made - which gives them an additional layer of intelligence and an edge over more traditional chatbots.

In addition, Contextual chatbots do not limit themselves to analysing the few last sentences of the conversation, but are capable of taking into consideration the context of the entire conversation history - often referred to as contextual memory. This allows Contextual chatbots to remember what has been said before, and therefore understand high-level abstract pronouns such as ‘it’, ‘this’, or ‘that’ when being used later down the conversation.

A prime example of the power of having contextual memory is provided in the following conversation, in which a customer is looking to buy a KIA Seltos. In a first question, the customer asks for the specifications of the car - thereby specifically mentioning the type the person is looking for. Next, the person asks a question that is entirely unrelated to the type of car he is looking for.

However, when later on in the conversation, the customer asks for the price of ‘this car’, the chatbot is capable of leveraging its contextual memory to infer that the customer is referring to the KIA Seltos from earlier in the conversation.

Kia Motors Contextual AI Chatbot Example

Another factor that sets Contextual chatbots apart from other types of chatbots is the fact that they are able to acquire world knowledge. Whereas this concept can be interpreted in a very broad way, world knowledge allows contextual chatbots to handle and interpret references to places, people, and objects, which - in normal circumstances - can only be understood given the context of a conversation.

A commonly used example to illustrate such contextual world knowledge is one that revolves around referencing things like sport teams or cars. When visiting a European sports betting website and asking a chatbot about “the result of the rangers’ game”, the chatbot’s contextual world knowledge allows it to infer that the person is referring to the score of the latest game that was played by the Rangers Football Club - which is a football club playing in the Scottish Premiership.

However, when visiting a car comparison website and asking for information about the Ranger - such a contextual chatbot is able to leverage its world knowledge to infer that you are talking about specifications of the Ford Ranger - which is a type of car built by American car manufacturer Ford - instead of the Irish Football Club.

Another example of leveraging world knowledge is a Contextual Chatbot’s capacity to connect to so-called Application Programming Interfaces (APIs) - which are software intermediaries that allow applications to talk with each other and exchange information in a computer-understandable way. This allows Contextual Chatbots to access a vast amount of information - including weather data or public transportation data - within fractions of a second.

A reduction in consumer switching costs have pushed businesses into investing in digital technologies that are capable of capturing and monetizing customer attention as much as possible. Irrespective of whether your website visitor is looking around for granular information, seeking a little after-purchase service, or just wandering around - contextual chatbots are bound to be a fruitful partner in providing customers with the automated granular and thoughtful responses you are looking for.To learn more about how your brand can build smarter Chatbot through contextual AI and provide a better customer experience, contact us today to book your demo.