What is Conversational AI?

AI has a variety of applications that are anything from super-human to super-machine. Analytics, big data and automation are often the first places the mind of a marketer, brand manager, or support leader goes when thinking about how they can apply AI to their work, but Conversational AI offers a whole new category of capabilities that business leaders need to consider when they serve their customers and stakeholders.

Conversational AI is any machine that a person can talk to. This could be a chatbot on a website or social messaging app, a voice assistant or voice-enabled device, or any other interactive messaging-enabled interface. These solutions allow people to ask questions, get opinions or recommendations, execute transactions, find support or otherwise achieve a context-dependent goal through conversation.

Conversational AI is any machine that a person can talk to, and is usually engaged with today through chatbots and voice assistants.

Does this replace human beings? No, not really – sorry to disappoint you (if you’re disappointed by that kind of thing). In reality, Conversational AI gives us help and expertise in a variety of places that we simply don’t have enough humans for, where humans will find the task mundane and discouraging, or where humans are too expensive to employ. In addition to that, when Conversational AI automates some of the more mundane activities humans have to do, it frees them up to be more productive.

What’s really compelling about all this is that Conversational AI represents a whole new category of engagement for brands – whether it’s in marketing, sales or support – and this engagement makes consumers feel closer to a brand, allows the brand to get more insights about their customers, and improve sales outcomes. Yep, that’s right, it grows revenue, whether you involve a human in the conversation or not.

 

What are the different versions of Conversational AI that a company can implement?

Chatbots

Chatbots were initially expected to catch fire in automating customer support, but the chatbot trend has evidently gone in a totally welcome-yet-unexpected direction. Contact centers are something incredibly difficult to automate, and only the largest of companies can generate any meaningful return from an AI investment by automating significant portions of their customer support.

In reality, chatbots are living in a whole host of places like websites, Facebook Messenger, iMessage, display advertising, and other channels in the future. In most of these circumstances they’re responding to more than just support questions – they’re actually allowing people to discover the products they like and want to buy.

This isn’t the only answer to the innumerable choices we have a consumers today, but it’s one of the better ones yet, because it actually allows people to have a conversation and think things through with the help of an expert assistant. Industry types have started calling this new category Conversational Marketing or Conversational Commerce, depending on the circumstance, but these are all things that are only possible with new developments in conversational AI.

In this case, nailcare brand Sally Hansen leverages a unique personality and tone of voice to drive the enthusiasm and engagement from its conversations with consumers.
In this case, nailcare brand Sally Hansen leverages a unique personality and tone of voice to drive the enthusiasm and engagement from its conversations with consumers.


Voice Assistants

Voice assistants are very similar to chatbots, but because people have to speak aloud to communicate with them, the market has seemed to evolve towards a variety of functions that are not necessarily transactional. For instance, top uses indicated by consumers are often things like making phone calls, playing music, setting alarms or reminders, answering questions, providing the weather or controlling home automation systems. Developers for brands or other organizations have the ability to publish skills that allow consumers to ask branded questions, have on-boarding experiences or otherwise get more brand guidance, recommendations and so on.

The critical consideration here is that brands can benefit from deploying to voice assistants, but voice assistants alone won’t drive full-funnel engagement. Conversational AI over voice has seen limited adoption for transactions, but it does offer a compelling entry point for consumers to begin product searches, ask consideration questions and lay out in their minds the aspects of particular products they’re looking to consider. All this is why conversational AI over voice should be one element of a multifaceted conversational AI brand strategy.

 

Mobile Assistants

Mobile assistants like Siri, Google Now and others, much the same as Home Voice Assistants like Amazon Alexa, Google Home and Apple HomePod, can work much the same way for brands, except in most cases, consumers are using mobile assistants to perform the functions that they need done quickly but when their hands are full. This would equate to text-to-speech functions while driving, sending quick messages, asking about the weather or for the results of a search engine query.

This offers another useful entry point for brands, and many brands are experimenting with things like mobile voice ordering. For most, however, a voice-only Conversational AI strategy again sets an unnecessary limitation on the degree to which customers will engage and re-engage with a brand. If they have to talk out loud to a brand at all times, they’re likely to avoid interactions in most circumstances – either where they want to message privately, or where there are simply other people around them.

 

Interactive Voice Recognition Systems

This is conversational AI-classic, the original systems that brands began leveraging to create service automation in their contact centers and reduce their overall customer service costs. Based on voice recognition technology that started to come to market in the 1990s, many consumers weren’t very big fans of systems like these, but companies couldn’t afford not to use them because of the cost savings they offered.

Now, they are being used in part, whole, or on balance with other customer service automation techniques. T-Mobile, for instance, recently announced that they would be doing away with the systems to add a more human presence to their customer service, but ultimately they would also incorporate new conversational AI like chatbots and obviously find other efficiencies within their customer service offering, as well as their overall customer experience.

 

I know Conversational AI when I see it, but what goes into creating it?

In order to understand the technology more deeply, it’s important to understand how a business can arrive at an effective use case for conversational AI. This requires understanding the existing business, setting a goal to improve and the channels to do it on, scoping a brand voice and finding the right expertise to implement the technology.

 

Understanding Your Business

Brand’s don’t need much to implement basic Conversational AI. For instance, there are many places that individuals can build their own chatbot. But in the case of leading brands, they need to consider both the overall strategy of where and when they would like to empower consumers to use Conversational AI, and then also consider what is required for that AI to contribute to an enhanced customer experience that improves both the adoption and retention a brand experiences.

Establishing this scope means you need to consider things like whether you want to start by improving your marketing, commerce or support efforts. This means considering things like what questions your customers have, what type of transactions you offer, whether your consumers have specialized needs in their choices between products, and how interactive you want your ecommerce or in-store experiences to be in total.

Chatbot use cases are the most variable, within and between industries, but more and more businesses are using them as revenue-generating Conversational AI. Source: UpWork.
Chatbot use cases are the most variable, within and between industries, but more and more businesses are using them as revenue-generating Conversational AI. Source: UpWork.

 

Setting a goal for what you want to improve

Conversational AI can improve a lot of things. At its base, it can improve your response rate to questions of all kinds, but many of the questions that consumers have which aren’t being addressed today relate to product consideration. Following suite, your conversational AI can then better engage consumers than a traditional web experience, and provide the instant answers that human operators might not necessarily be able to (or would simply be too expensive to deploy).

Automating this sort of one-to-one engagement empowers your brand to begin measuring things like engagement not in clicks, likes or shares, but in actual time devoted to a two-way interaction. Conversational AI attracts consumers into consideration conversations that help them make purchase decisions, allowing measure your consumer engagement down to the minute and up to the moment a consumer makes a purchase or repurchase from your brand.

Conversations progress in short increments, but that allows Conversational AI to capture relevant insights on consumers as they get guidance and recommendations.
Conversations progress in short increments, but that allows Conversational AI to capture relevant insights on consumers as they get guidance and recommendations.

Aside from that, there are other aspects of your funnel that conversational AI can improve. Those conversations, for instance, can help generate unique customer insights. Looking to generate insights on a billion consumers like Unilever is? Conversational AI may help in that endeavor.

Conversational AI can also allow you to have a direct impact on increasing sales and customer satisfaction, whether its average revenue per user or overall customer satisfaction score. These are all ways in which a brand can consider using conversational AI to compensate for its weak points, emphasize its strengths, or both.

 

Deciding on the channels that best suit your business

Conversational AI is not a one-and-done kind of solution. Chatbots, voice skills and other implementations may be considered novelties by many, but the underlying benefits they generate have caused many brands to not only let them “stick around”. Many brands are now looking for ways to expand their efforts and find new channels and use cases in which conversational AI can continue yielding accelerated returns.

The most significant cases to consider in this case is what “modalities” you want your conversational AI involved in, and what it can do best in each of those domains. Advertising, marketing, commerce, sales, support and general assistance are all applicable disciplines for an AI, but in advertising for example, they might work best to call attention to an offering. In marketing, they might work best to guide consumers on a website. In sales or commerce, they might work best to give recommendations or advice to consumers as they approach a final purchase.

The applications go on, so it’s important for brands to consider where consumers need help and answers most of all – it may not be as obvious as you think. Our greatest variety of unanswered questions comes before we buy something. Brands that win out this evaluation period with AI are much more likely to win out overall.

From ads to search, social media, messaging apps and voice assistants, brands have multiple channels to pick from for their AI deployment - or can use them all at once.
From ads to search, social media, messaging apps and voice assistants, brands have multiple channels to pick from for their AI deployment – or can use them all at once.

 

Examining Your Brand Voice

Each brand, like each person, communicates with their own distinct voice and attitudes. These attributes not only allow a brand to be remembered, but they also improve a brand’s connection with the consumers who are already most likely to become their advocates or champions. Brand voice is critical to determining how a brand will tap into new markets of consumers, and with conversational AI, a brand’s voice comes to the fore because for the first time brands can actually talk with consumers.

So, when considering your voice, think about how you want to make your consumers feel. Is it fun? Is it cool? Is it loved and cared for? Most usually, it won’t be the opposite of those things, but each brand can carry this forward in its own key and unique ways. For instance, a beauty or skincare brand can use a lot of emojis and images, while a bank or a telecom can use lighter symbolism and an enthusiastic but straightforward voice. An edgier consumer brand may be willing to take conversations down an unexpected path in order to create a more memorable experience, while staple services will want to be dependable and predictable, but at the same time take the opportunities they can to stand out and reassure the consumer.


Seeking Out Expertise

Ultimately, new and emerging technology always has a few early providers and adopters who set the stage for what works and what doesn’t. These unique leaders stake out new territory and take on the risk that’s required to find out what will work for the rest of us. In most cases, the type of conversational AI a brand will see the most benefit from needs to incorporate Natural Language Understanding, based on deep learning and other recent developments in AI that create the most satisfying conversational experiences possible.

In this case, maybe your brand wants to take on that risk and do everything itself. If not, or if you’d like to realize the full potential and value of Conversational AI as soon as possible, it’s important to consider what experts – designers, technologists, vendors and consultants – can give you the quickest insight into your own business. Seeking them out will allow you to figure out the best way for you to realize your vision for Conversational AI, without bearing the costs so many others have before.

Deep learning isn't the only technology that has enabled the recent explosion in Conversational AI, but mixing it into existing conversational experiences greatly enhances the amount of inputs they can adequately address. Source: House of Bots.
Deep learning isn’t the only technology that has enabled the recent explosion in Conversational AI, but mixing it into existing conversational experiences greatly enhances the amount of inputs they can adequately address. Source: House of Bots.

 

How can a brand benefit from Conversational AI?

The benefits of Conversational AI are multiple. Firstly, it not only enables brands to offer instant answers about a variety of consideration questions consumers may have, but it ultimately enables brands to generate more minutes of engagement, consumer insights and sales as a result. Conversational AI, because of the quick answers it provides, can also positively benefit customer satisfaction.

Each industry sees different outcomes, but in a world where mass marketing is slowly giving way to one-to-one brand building, conversational experiences will play a key role in generating these one-to-one experiences and leveraging the information they provide, giving brands an edge no matter their core offering.

 

To get started with Conversational AI, consider contacting Automat. Automat’s Proprietary AI Technology, leading Conversational Cloud Platform and expert Customer Delivery and Success team provide brands with all the resources they need to better engage, connect with and convert consumers.

1