Voice Assistants (VAs) and chatbots have now earned a spot in the tech tool stack of many hotels but choosing the right solution for your hotel is not easy. If you are planning to invest in guest-facing conversational technology, but every option feels the same, this article is for you.
First, the choice is not between voice assistants and chatbots. Voice or chat are the different ways guests interact with the technology, also called conversational user interface (conversational UI). When you’re evaluating which solution will best suit your hotel, you should base your decision on the technology behind it. The interface, which is voice or chat, comes second.
There are 2 main types of conversational agents used in most solutions on the market:
Each is there to meet different goals for hoteliers and each fits a different application in the industry. After reading this article you’ll understand:
Let’s talk basics!
Rule-based dialog agents work with questions, structures, answers, and keywords that are predefined by the hotel staff. Because they’re implemented on specific sets of rules, their scope depends on the bot designer.
Rule-based dialog agents are usually Q&A-based flowcharts, therefore, not categorized as AI technology, but more of a digital self-service system. (We’re talking about the same technology that powers automated telephone response or IVR.)
User Experience: Solutions with rule-based dialog agents are also called Rule-based chatbots, decision tree bots, or scripted bots.
The interaction with a rule-based conversational technology can be a simple Q&A flow, or a menu-based interaction (decision tree), where guests choose from a list of options.
Rule-based dialog agents follow predefined rules, their responses are based on keywords, and there is no intelligent algorithm looking for context.
Cons: Rule-based dialog agents are implemented on specific keywords, can’t answer questions outside predefined rules, or understand typos, which can be frustrating for the users and limit their freedom of expression. Unlike conversational AI, rule-based dialog agents don’t leverage the data from interactions.
Hoteliers must continue maintaining and making new keyword rules to make these bots useful. That’s why most hotels usually stop using them after a few months.
Price: Rule-based dialog agents are simpler, with no training engine they need less building effort than Conversational AI, therefore they are usually more affordable.
If you don’t have many queries from the guests, you want to implement a digital FAQ instead of printed information in the guest room, but you don’t expect much in return.
For your hotel, this means that the staff designs the questions and responses first. Guests go through a pre-designed Q&A tree that answers simple queries like check-out time, gym opening, or Wi-Fi password.
By pre-defining the structures and answers, you can control the entire conversation, but you miss an opportunity to learn about what your guests really want to ask.
Conversational AI is not based on keywords, but it tries to understand the intent of the guest, and what the guest wants. The accuracy level depends on the number of preset *samples (utterances) of what guests could ask for a specific intent and all the data is trained up to date by the AI model.
Guests can ask the same question in different ways:
‘Set alarm for 7’
‘Wake me up at 7’
‘Morning call for 7’
User experience: Conversational AI understands unscripted queries coming from the guests, even with grammatical mistakes or spoken with different accents (this is handled by ASR, automatic speech recognition, layer). Conversational AI can also learn to understand multiple intents in a sequence. This is why interactions with conversational AI feel more natural and human-like compared to rule-based dialog agents.
For people, voice is so far the easiest and the most natural way to interact with technology. Adding voice interface to conversational AI levels up interactions with technology and makes the conversations feel more human-like. That’s why Voice Assistants have much higher usage compared to the web-based solutions powered by rule-based dialog agents.
On average Voice Assistants answer every in-room query in 3 seconds with a 90%+ success rate while the same task usually takes over 3 mins for web-based chatbots. (Guests scan the QR code > download the App or open the browser > chat their way to answer the question)
Cons: Conversational AI requires a dedicated dataset for training and implementation, and therefore a longer time to develop.
Price: Because of its advanced technology and training, the price of solutions powered by Conversational AI is typically higher but also leads to a much better task-completion rate, higher usage and better guest experience.
For your hotel, this means that guests can ask questions freely, and you get more insights into their real needs. Conversational AI continuously keeps improving without much further assistance. The more it interacts with guests, the better it will become at understanding the intent, and at answering guest requests.
In this article, you’ve learned the difference between conversational AI and rule-based dialog agents. With this information, you should be able to make a well-informed choice of solution that best matches your business goals.
Rule-based bots and conversational AI each serve a different purpose in the industry. If you are tired of printouts and want to give your guests a digital form of hotel FAQ, rule-based bots might be for you. But if your staff has a large volume of conversations with the guests every day, and if you’re looking to turn this data into business knowledge, opt for conversational AI.
As exciting as new technology is, it’s necessary to truly understand it before you make the purchase. AI system is a long-term investment, it comes with benefits that go beyond the shiny guest experience. It helps your staff free up more time, work more efficiently as a team, and centralize requests and interactions in one place, minus the fuss.
Do you want to learn more about what we’re building? Contact us to schedule a demo.