As IVA technology becomes more and more a part of everyday business, it is important to be aware of what Intelligent Virtual Assistants can and can’t do today. It is also wise to keep an eye towards the future and plan for what IVAs may be able to do a few years down the road.
What Are Some Common Misconceptions About IVA Today?
One myth floating around out there is that IVA can be “tricked” or easily broken. It is true that it can be fooled (i.e., it’s not going to understand something that is outside of its context or domain). For example, a customer asking a question where the answer is unknown or unanticipated is common. At this point in time it would be irresponsible to have an IVA give an answer that it was not programmed to offer. If a patient asks a doctor’s IVA if they should take aspirin for an ailment, it would not make sense (and possibly unethical) to have it answer (at this point IVAs would not have all the knowledge a doctor would have regarding diagnosing patients, not to mention that it may not have access to medical history where certain recommendations could actually be harmful).
That is why it is critical that in-depth discovery and mapping of the potential conversations is prioritized. It will be time well spent. Programming happens when all the human elements have been defined. The “defining” is the hard part. The programming is the easy part.
Others believe that the IVA is impersonal – a robot-like entity that is not as “warm and fuzzy” as a real person. For certain instances this is true – think mental health issues or life and death medical issues. That is why it should never fully replace human customer service and interactions. But it should, however, augment customer service when a human is not available, like after hours or when all employees are busy. A well-designed IVA’s convenience outweighs the lack of personality (and in some cases, an IVA may offer more personality than some humans).
Depending on the use of the IVA, personality can be added into the technology. In light hearted situations, ‘Facts of the Day’ or even dad jokes can be programmed into the script. Or adding pauses at certain points, or even adding fake sneezes can bring an IVA to life.
Lastly, some people confuse intelligent IVAs with plain-old bots/chatbots. A bot lacks context and has a very finite number of responses it can give. After an initial query and answer, a basic web bot won’t do much beyond telling you to check back tomorrow, or taking a message. An IVA is trained to perform certain tasks that go well beyond chit-chat, and provide lots of value when none existed before. For instance, if a user is asking the question ‘How do I reset my password?’ A chatbot may refer them to an article explaining how to do the process. An IVA will ask for the user email address and send a password reset link.
What About Developing An IVA?
In order to build a robust, useful IVA for your business, a developer needs to know conversational logic, your particular customer experience, and the best ways to funnel many different customer interactions to completion. An IVA that is going to be on the front lines for your business must be designed with the entire anticipated interaction in mind. This requires knowledge of your customer’s business processes, and their customers’ intent. This intent is the reason for their contact with you – and each reason has a specific path to follow. While customers may have 25 possible intents, you might need 25 responses or ways of handling each of those intents.
The IVA solution must mirror your business processes, mission, purpose, and vision. That’s why consulting with experts in the IVA field is always recommended. From planning and development to testing and implementation, so much needs to be considered in order to build the best IVA for your customers and your business.
What Does The Future Of Intelligent Virtual Assistants Look Like?
To the delight of many, the emotional analysis ability of IVAs will continue to improve. This will allow them to adjust their responses based on the user’s mood and sense of urgency.
In the future, speech recognition technology will become more advanced. Your voice will be catalogued and combined with other data points to get a clearer picture of you as a person, a consumer, and a citizen. Voice verification currently exists and there is no real barrier (outside of privacy issues) beyond the sheer amount of data and processing required to map a voice to a person, similar to facial recognition. Everything could be voice-driven. Think of an ATM at your bank. It could be equipped so that it looks at your face and matches it with your voice – no PIN required. And to thwart would be criminals, a secret word or wink of the eye could trigger an alarm to the police that you’re taking out money against your will.
IVA will also continue to improve intent detection (i.e., what does this person really want), and make more decisions “on its own” about what to say next. Think of a case when someone is calling for store hours 10 minutes before closing. What the caller is really wanting is for the store to be open in 15 minutes when he finally gets there to pick up a birthday gift. For these situations the IVA can be programmed to give store hours, but also alert a store representative of the question. If there is flexibility in closing times, the rep can pick up the phone and talk to the customer to let them know they would be happy to stay open an additional 10 minutes to accommodate them.
Do you still have questions you aren’t sure of when it comes to what an Intelligent Virtual Assistant can and can’t do? Is there a project where you feel an IVA would be a great solution, but you just can’t figure out how it would be developed? Give us a call. Our experts love talking with businesses to figure out how the latest technology can help streamline customer experiences and make businesses more efficient.