Conversational AI platforms and Conversational interfaces are still in their infancy. As with all new technologies, it’s important to analyze them with the hype cycle in mind. Gartner placed them last year at the very early stage, just before the peak of inflated expectations. This means that we have a couple of years ahead of us until we come to a point of real productivity and usefulness.
Having this in mind, let’s speculate and dive into the most important trends that might appear in 2019 in this dynamically evolving area.
The growth in this space is one of the most often mentioned predictions. Voice assistants and smart speakers are coming out of our kitchens and crawling into our cars and offices. As voice assistants like Google Assistant and Amazons’ Alexa add more features and functionalities, the smart speaker devices will start reaching mainstream market. One trend to watch out for here is multi-modal interfaces, which blend voice with touch and displays to bring enhanced experience.
Bots are helpful in every cycle of e-commerce sales – both pre-sales as well as post-sales. Merchants currently use bots for customer acquisition, retargeting and customer service. Shopify stores have seen anywhere between 3% to 25% increases in revenue by adding chatbot. They look and work great on mobile devices and can help users focus on one task at hand by hiding conventional navigation and webpage elements. As business users will start asking for ROI, bot projects will be pushed further to generate real revenue.
Conversational User Experience
We have a lot of guidelines for conventional web usability and web/mobile interfaces. As conversational interfaces gain traction, we will need a new set of guidelines – perhaps giving birth to a new role – Conversational User Experience experts. People who blend technical and psychological knowledge with ability to analyze large amounts of analytical data to improve the conversations.
Forming of ecosystems
To get to Level 5 of conversational systems, bots will need to teach to talk to each other. In order for this to happen, we need to develop a common set of rules and interfaces for AI-to-AI communication. We should start see this approach appear in major platforms, who will start exhibiting services in fashion similar to traditional REST API’s or GraphQL interfaces, but for skills.
As the user conversation and focus moves to bots, so the ads need to follow. I predict that in 2019 we will see first adservers aimed at bots, using the context of conversation to serve relevant commercial information. As those products become available, we will probably see a new breed of RTB (Real Time Bidding) platforms that will allow brands to hop into conversations in real time with (hopefully!) relevant content.
Multi-modal Voice Assistants
Studies show that Voice Assistants prove the most useful in places where our hands are busy. For majority of users this means kitchen, car and… bathroom. Those use-cases benefit greatly from adding a display. Think asking for a recipe in kitchen and seeing a walk-trough of how to prepare the meal or monitoring the baby as it sleeps in his bed. As smart speakers turn into multi-modal devices, manufacturers start adding person-to-person communication features, turning them into a modern version of walkie-talkies.