What's the difference between chatbots and virtual assistants?
Forrester's Thomas Husson gives his take on the difference between virtual assistants and chatbots. He says one is closer to a 'dynamic FAQ' and the other is a third-party aggregator.
There's always been a level of ambiguity around the terms chatbot and virtual assistant -- an ambiguity that IBM Watson Vice President and CTO Rob High tried to clear up in a recent Q&A titled, "Who's talking? Conversational agent vs. chatbot vs. virtual assistant".
According to High, there are subtle but distinct differences between the terms and technologies. These differences have to do with the extent to which chatbots and virtual assistants are able to engage the user.
In this Ask the Expert, Thomas Husson, a Forrester Research vice president and analyst whose research focuses in part on conversational interfaces, has his own take on the differences in terms.
Husson explains why he considers chatbots to be closer to "dynamic FAQs" and virtual assistants to represent the aggregation of many chatbot and personal assistant experiences. Read his explanation to find out which conversational interface has the advantage in the consumer market and which is better suited for the enterprise.
What's the difference between chatbots and virtual assistants?
Thomas Husson: Chatbots help users complete tasks through simulated conversation. Depending on the individual chatbot's capabilities, it will determine user intent and then respond using voice, text, photos, web content and emojis. Users respond through voice, text or by selecting an on-screen option.
This is considered a conversational interface. Most chatbots use text for interactions and they are prevalent on messaging apps, but nothing prevents consumers and enterprises from utilizing voice in a chatbot conversation.
Thomas Husson
Virtual assistants -- also called intelligent assistants -- orchestrate agents or services from third parties on behalf of consumers. Bots are one form of an agent. Virtual assistants rely on context (e.g., user input, localization capabilities and access to information from a variety of databases) to continue to refine the quality of responses to a user's requests. These assistants are essentially guessing, but they get better over time as they're used more. Examples of intelligent assistants include Amazon's Alexa, Google Assistant and Apple's Siri.
Smart speakers
There is a lot of emphasis these days on smart home speakers like Amazon Echo or Google Home. The buzz has continued this year with the launch of Apple's HomePod and others to come. This is indeed a fast-growing category, but what matters is not so much the new device category itself, but rather the assistant powering it. The better the assistant is at leveraging machine learning and AI-related technologies, the smarter it will be and the more contextual the experience will be. At the end of the day, these assistants aim to understand your context and anticipate your needs.
Enterprise vs. consumer market
Chatbots today -- particularly in the U.S. -- are often used for customer care services and they can definitely let consumers and employees accomplish tasks faster. Chatbots help users navigate complex FAQ lists or portals, but they can also be used for broader marketing purposes and help at each phase of the customer lifecycle.
Due to the reach of messaging apps like Messenger, WhatsApp, WeChat and Line, these will shape the first experiences that people will have with machine learning conversational interfaces. However, today, most chatbot experiences do not deliver on their promise and can be quite frustrating.
Virtual assistants -- especially Google Assistant and Apple's Siri -- have an advantage in the sense that they are and will be embedded into many smartphones, creating scale and usage beyond smart home speakers and other connected objects. The value here will come from the ability to deliver new contextual experiences and to save customers time.
Capitalizing on chatbots and virtual assistants
A chatbot should start with a specific use case before fully leveraging AI. It can drive value if you take the time to train it. The same is obviously true for a virtual assistant since it is also based on machine learning and AI. However, the objective is much more ambitious for these intelligent assistants because the digital giants developing them [e.g., Apple, Google, Amazon] aim to aggregate many chatbot and assistant experiences.
Theoretically, both chatbots and virtual assistants can be considered conversational interfaces. However, this is not today's reality. Most chatbots are closer to dynamic FAQs than to true conversations -- and that's fine. Companies need to start with a simple use case like saving consumers time and helping them find what they're looking for, and chatbots can do that.
Thanks to machine learning and other AI-related technology advances, especially natural language generation (NLG), chatbots will eventually produce relevant content easy for human beings to understand and interact with. Today, the focus is on understanding user intent (what the consumer is asking the bot via natural language processing and natural language understanding) and less on generating a true conversation via meaningful content and NLG.
I don't think it is necessarily a good thing to think [of chatbots and virtual assistants] as one versus the other. To make the most of these interfaces, [enterprise] brands will have to deliver their promise by enabling true two-way conversations and must get ready to activate their chatbot's personality -- via things like voice tone, charisma, etc. -- on these new interfaces.
