What’s The Difference Between Conversational Chatbot Solutions, Rules Based Chatbots, and Traditional AI?

Written by Can Ozdoruk on Apr 23, 2020

The History of Chatbots

As you may already know, chatbots are software that use natural language processing (NLP) to engage in conversations with users. But that doesn’t mean that all types of chatbots are created equal. Below, we are going to demystify three common terms for chatbot that you may be hearing across the industry: conversational chatbots, rules-based chatbots, and AI. 

You can include these bots in mobile applications, messaging apps, websites, email, and even voice platforms like Alexa. Online retailers are integrating their chatbots with Shopify to increase revenue. 

Along with countless benefits, many companies use chatbots for customer service as a way to provide immediate resolutions to common issues. Conversational chatbot solutions and artificial intelligence have never been more popular than they are today.

In fact, data from Google Trends shows that interest in chatbot solutions has increased ten-fold over the last 5 years.

Interest in chatbots over time, from January 1, 2004 through October 2020, according to Google Trends

During this explosion of interest, “chatbot” has evolved into an umbrella term that may inaccurately describe what a chatbot can and cannot do. Chatbots and conversational AI technology are often used interchangeably. In reality, the capabilities between chatbot technology and artificial intelligence are very different. We’ll explore more about what separates some chatbots from others below.

Chatbots vs. Conversational AI. What exactly is the difference?

It’s important to understand why modern artificial intelligence chatbots (also known as Conversational AI or AI agents) differ greatly from first-generation (rule-based) chatbots. The first chatbots adopted by companies were based on stringent rules and rigid decision trees that often led to frustrating user experiences. On the other hand, modern chatbots are more forgiving when it comes to following strict rules, enabling users to engage naturally in conversation. This evolution from rules-based chatbots to conversational chatbots is a huge factor in the explosive growth of interest in chatbot technologies.

More companies are looking to virtual assistants and conversational interfaces to provide anytime, anywhere customer support. So, it’s important to have a clear understanding of different technologies. That’s because the scope of a conversational agent initiative and the end-user experience is vastly different. Rules-based chatbots are limited to very basic scenarios. On the other hand, AI-powered virtual assistants are capable of engaging in natural language understanding, participating in 1:1 conversations due to machine learning, deep learning, and conversational experience.

Companies must ensure that they are adopting the right technology for their business and their customers. This is because the customer experience plays a critical role in consumer buying decisions and loyalty. Here is an example between modern conversational AI and basic bot technology:

If a person asks a question that a chatbot has not explicitly been trained to handle, it is easily confused. Conversational virtual assistants enable users to engage in natural, human-like conversation. This is because conversational chatbots are better able to understand intent, linguistic variance, and complex phrasing.

What Are Rules-Based Chatbots?

Rules-based chatbots can automate customer service in very specific scenarios. For example, looking up an order status or browsing through a product catalog. Basic chatbot technology moves the conversation forward via bot-prompted keywords or UX features like Facebook Messenger’s suggested responses. (As compared to typing in a question in free-form, using slang and engaging naturally in a conversation). 

Basic chatbot platforms have limited, if any, natural language processing. Typically, the bot will ask a user a question and display a few responses in which a person can select from or it will identify a specific keyword in a user’s question. Based on a person’s input, the conversation moves forward on a specific path. With pattern-based bots, what a user says must explicitly match with how a bot was pre-trained in order for it to understand and move the conversation forward.

In regards to this, variations of a question must be pre-trained for a chatbot to accurately understand what a person is trying to say. For instance, a virtual assistant is trained to understand “Where’s my order?” If a customer asks the same question slightly differently, “Is my package arriving today?”, the bot will not accurately understand the intent of the question is “order status.”

That is, unless it has been explicitly trained to do so within the labeling and learning provided in its training data.

Rules-Based Bots And The User Experience 

Chatbots lack semantics and advanced Natural Language Processing to understand the context of a message.

The user experience with rules-based bots is often alinear. If a person says something that is not preempted, a chatbot will get confused. The virtual assistant will most likely repeat the same question until it understands a response. For example, a chatbot designed to help people order a pizza will not know how to respond to a customer asking for nutritional facts as they are selecting toppings. 

How to Train Rules-Based Bots

Chatbot training is a manual process and requires programming every flow and utterance of a question. A human workforce also identifies and implements ongoing improvements. 

If you’re deploying a rules-based bot, make sure that you select a very specific use case. Fandango, for instance, has a bot that asks people for their zip code and pulls up movies playing locally. In another example, The Wall Street Journal lets users type in a stock symbol to get live stock quotes. These use cases are very specific and well defined and work well for bots. 

Be upfront with your customers on a chatbot’s capabilities. You’ll need to provide an alternative method of getting support for other matters (i.e. I’m the Order Tracking bot. To find your order, type in your confirmation number below. If you need something else, please call ….). 

Conversational Chatbots

Conversational chatbot solutions are AI-powered virtual agents that provide a more human-like experience. In opposition to rules-based chatbots, they are capable of:

  • carrying on a natural conversation
  • understanding the meanings of words
  • understanding misspellings
  • continuously improving over time

Because of these important differentiating features, conversational chatbots provide a greater user experience through the use of natural language processing and leverage semantics to understand the context of what a person is saying.

Conversational Chatbot Examples

Here’s a quick example scenario of how conversational AI works: “I got the side table delivered yesterday but it looks like it might have been broken while in route. There’s a crack in the front. Can you help me? I would like my money back.”  An AI-powered virtual agent would be able to decipher that a person is looking to return an item and receive a refund.  An AI thinks like a human, not a robot, and is able to maintain a conversational flow.

AI-based conversational chatbots use machine learning technologies to understand, contextualize, and predict to accurately respond to user inputs. They enable companies to provide hyper-relevant personalized engagement, not generalized support. This can be done by training algorithms used in these chatbots with historical data from real user responses and can be optimized with ongoing user feedback (reinforcement learning). Like humans, AI virtual agents are able to decide the next best action based on a variety of things including contextual-factors, customer profiles, sentiment, or business policies. Furthermore, it can alter how it responds based on real-time sentiment analysis. For instance, an AI Agent treats a person who checks the status of their (on-time) flight differently based on how they react. A virtual agent would presume that a person who responds with “Oh no!!”  that they are likely to miss their flight. 

Two Different Types of Conversational Chatbots: Generative vs Retrieval

Continuing, there are two subclasses of learning-based chatbots: generative chatbots and retrieval chatbots. Generative chatbots can dynamically create responses in real-time, and retrieval chatbots select from a pool of responses based on the person’s message to the bot.

AI-based conversational chatbots leverage semantics to understand the context of what a person is saying. Therefore, these bots can engage more naturally in conversation, and respond to more inputs without being explicitly trained on every single way a person might phrase their question, like the flight example above. Traditionally, these bots may not have been as accurate as pattern-based methods and used to take a long time to train. However, there are a few companies, like Netomi, that have built robust NLP engines that accurately understand user inputs up to 95% of the time, which means scaling and training are now exponentially easier and the end-user experience is much better than pattern-based bots.

Conversational chatbot solutions powered by AI also support multi-turn dialogue. This is the ability to switch between various user questions within a single conversation. This is what sets apart a human-like AI versus building chatbots. An AI-powered virtual agent responds without getting confused if a person pivots the conversation. For instance, a person can ask about the price of checking a bag in the midst of checking flight status. In conclusion, AI can also understand more short-form and slang than chatbots, giving conversational chatbots a wider range of use cases than rules-based chatbots. 

How to Train Conversational Chatbot Solutions

AI training is a combination of supervised and unsupervised learning. AI can learn from historic data. With customer service, this includes customer support email, chat and messaging logs, to identify and group together similar questions and scenarios. Training is on auto-pilot. An AI learns how a situation has been handled and teaches itself to act in the same way. 

AI also uses deep reinforcement learning to improve over-time based on real-life interactions. AI-powered virtual agents are able to determine patterns based on how end users are responding in various circumstances. This is based on things like customer segmentation and contextual factors. For instance, if meal-delivery customers have issues with changing their subscription day, an AI would learn to proactively offer this information. 

The richness of the technology has Gartner predicting that by 2021, 15% of all customer service interactions will be completely handled by AI1. The best AI chatbots tend to be the most self-sufficient when it comes to adapting. When you hear about terrible chatbot fails, those are likely stemming from less-sophisticated bots and/or an improper way to set them up – basically launching a bot without enough training.

Key Trends in Conversational Chatbot Technology 

There are a few emerging trends that are propelling the sharp rise in the adoption of conversational chatbots. Take a look at these key chatbot trends:

  • Personalization – personalization involves chatbots tailoring the interaction based on customer profile and behavior. For example, an AI bot could provide a hyper-relevant cross-sell recommendation by learning that a customer prefers certain brands or types of products. By incorporating customer experience personalization, chatbots respond on an individual level, providing more meaningful interactions.
  • Voice recognition – voice recognition enables faster, hands-free interactions for users, making AI bots even more convenient. Examples of voice recognition can be found in the array of personal assistants, including Google Assistant, Siri and Alexa. Companies, including WestJet, are also launching skills on voice platforms to provide yet even more choice with how customers receive support.
  • Machine learning operations (MLOps) MLOps is a strategy used to automate and operationalize machine learning workflows. This strategy plays a role in chatbots by improving the speed and ease with which bots can be trained and improved. With such automation, bots are ready for market faster and can be more frequent, and easily updated.
  • Memory and context – many brands store customer information in customer relationship management systems (CRMs). When integrated into chatbots, CRMs can provide valuable information that enables chatbots to continue previous conversations with customers or look up specific details about the user. While this is often limited to profile details for privacy, chatbot engineers are working on ways to make queries more secure to enable broader interactions.

What To Keep In Mind As The Differences Between Chatbots vs. Conversational Chatbots With AI

Conversational chatbot and AI adoption is skyrocketing. In fact, according to Accenture, 60% of surveyed executives plan to implement conversational bots for after-sales, customer service, and social media. Accenture isn’t the only organization projecting big movement within the chatbot space – just take a look at these very telling chatbot statistics. At first glance, chatbot technology and AI-powered conversational interfaces appear very similar. When you go below the surface, though, the technology could not be more different. The initial training, the ongoing improvement, and the end-customer experience are not even close to being in the same league. 

Interested in learning more about artificial intelligence and chatbot technology? We’d love to discuss how our powerful AI chatbot platform provides the frustration-free experience your customers expect. Don’t use a robotic, limited chatbot solution that plummets your CSAT. Let’s chat.