If you’ve kept an eye on tech news over the past month or so, then you already know that AI-driven technology is no longer niche and experimental.
ChatGPT and Stable Diffusion have been some of the most prominent instances of highly impressive AI tech, but really, all specific iterations of AI technology represent a small portion of the potential of the category as a whole.
Perhaps the most important takeaway from these recent successes is that AI is here to stay.
But we’re not here to look at AI in a broad sense, but rather to focus on a specific AI use case: conversational marketing.
An expert on the subject, Alena Arkharova, has joined us for a closer look into the real-world benefits of conversational marketing and its potential for advancement.
Arkharova has shared her marketing expertise with a number of brands over the course of her career, including KPMG, Emerging Travel Group, and, most recently, Satisfi Labs here in the US, where one of her many accomplishments has been supporting the development of AI Chat for the purpose of conversational marketing.
But we should really nail down a definition of conversational marketing before diving into the nitty gritty.
What is conversational marketing?
“Conversational marketing is a way to engage with potential customers through chat and messaging platforms in a personalized and conversational way. The goal is to create a natural, human-like interaction that helps to build trust and establish a relationship with the customer.”
That’s how Arkharova describes the concept, and she elaborated by saying that conversational marketing can potentially include website chatbots, text messaging campaigns, as well as social media messaging.
AI technology tends to carry an association with inhuman qualities, or at least it has in the past, but modern AI tech emphasizes a more human impression.
Specifically, AI text generation has improved dramatically in recent years, learning from mass amounts of text written by humans.
Customer service chatbots have been around for years now. You’ve probably encountered these bots at least once, even if only as an introductory conversation to support from a human representative.
One of the current goals is to have these bots come across as even more natural, not to trick customers into thinking they’re speaking to a representative, but to provide fast, efficient support. In fact, customers might be more likely to ask certain questions to chatbot as compared to a human representative, but we’ll get to that later.
Here at the start, the most important point is that conversational marketing offers substantial benefits to both business owners and their customers.
“Conversational marketing aims to provide a more human and authentic way for businesses to connect with their customers, resulting in increased engagement, sales, and customer loyalty. Conversational marketing provides an immediate and personalized response, which improves customer engagement. It can also generate leads by providing an easy way for customers to ask questions and learn more about a product or service.”
Satisfi Labs’ AI Chat
Arkharova’s firsthand experience with conversational marketing has been Satisfi Labs’ AI chat project.
“Satisfi Labs develops an AI-powered Chat that our clients from the sports, entertainment, and tourism industries can place on their website, in the app, and on third-party messaging channels like Facebook Messenger to answer customer questions 24/7 and perform transactions.”
This chat can be easily adapted to fit the client’s needs and offer support relevant to the client’s line of products or services.
Built on a Natural Language Processing (NLP) model, this chat is able to analyze the context of a question and returns a response suitable to that context.
Arkharova posed an example scenario where a customer might ask a specific question about how to find parking. If the customer follows this with a more vague question, such as “How much will it cost?” the chat is able to look back at the previous question, determine the context, and provide the appropriate answer.
This hasn’t been possible with previous versions of CS chatbots, which we can speak to based on personal experience. These older bots tended to only respond accurately to highly specific questions, which led to customers having to guess which phrases and questions the bot would actually understand. This didn’t exactly make for natural and intuitive interactions.
But that’s changed. These bots are just better, and even more importantly, they continue to improve because they’ve been designed to continue learning from other conversations across the network.
“The system continuously learns and updates from shared knowledge across over 400 clients, which we call the ‘network effect.’ Each client benefits from industry contributors before them and adds to the intelligence and training today and in the future. As we say, our products are built for the industry and by the industry.”
Now let’s look at how this system can help guide customers through the purchase process.
Guiding through the buy
“Customers now expect answers on-demand to help make real-time purchase decisions. By providing personalized and engaging interaction, conversational marketing can help to increase sales by guiding customers through the buying process and addressing any concerns they may have.”
For many customers, research is an important part of the purchase process. Especially hesitant buyers want to learn as much as they can about the product before finalizing the purchase, and this is an area where AI chat can be extremely helpful.
But Arkharova also explained that Satisfi’s AI chat also offers integration with several services, such as Ticketmaster, Ingreso, and Tickets.com. What this means in action is that if a customer is already speaking with the AI chat and decides they’d like to go through with the purchase, that purchase can be handled via the chat.
Another crucial benefit of AI chat versus person-to-person support conversations is that AI chat is capable of working with a much larger number of customers simultaneously, which is especially important in the sports and entertainment space during large events.
Even a large staff would be unable to help such a large number of customers quickly. Instead, the AI chat handles simple questions, leaving the staff free to handle more in-depth conversations and support scenarios.
And even after an event or a purchase, the AI chat offers yet another advantage: tracking and analysis.
“Conversations can be tracked and analyzed, allowing businesses to gain valuable insights into customer preferences and pain points. Imagine you have direct access to every interaction of the users with the chat. You get real-time data on what your customers want and need, allowing you to adjust your marketing communication and improve the website content, product, or service much faster.”
The benefits are undoubtedly impressive, but the tech here still isn’t quite perfect, and it’s important to recognize the limitations in order to move forward.
The limits, and how to overcome them
The training process of the AI chat system covers as much relevant information as possible, covering what the business has already identified as common questions and concerns.
But there’s always a chance that a customer will ask about something that the AI hasn’t been trained on.
“Let’s say the client is a college athletics program. Suppose the customer decides to ask a question about academics, like an academic calendar or certificate programs. In that case, it won’t be in the scope of knowledge our AI is trained to respond to for this particular client. But if they ask a question about a sports schedule, it will be the scope AI can easily handle.”
Following this interaction, the problem could be mitigated, but this is indeed a limitation.
Further, Arkharova pointed out that the AI chat system is also specialized for the industries they work in: sports, entertainment, and tourism. But that’s by design.
“Our knowledge in these industries is one of the most advanced in the market, as we have been training AI for almost seven years. We want to keep providing high-quality service to our clients, so we will never launch a new industry that would be contextually far from the knowledge we’ve already trained.”
The goal is to provide a chat that’s been given a great deal of expertise across these particular industries in order to serve clients in those industries. This means that, rather than becoming an everything-AI, the chat system will continue to improve in this category to best serve brands and customers in that category. In other words, certain limitations are established specifically to keep the AI useful to a specific group of clients, and over time, it will become even more knowledgeable on these subjects.
One of the most interesting points brought up during our conversation with Arkharova was the idea that customers tend to interact differently with a chatbot than they would with a support representative. In many cases, that’s actually an advantage.
“There is evidence to suggest that customers may be more likely to ask specific questions when speaking with AI Chat or a chatbot. Customers are more likely to ask simple, straightforward questions when talking to AI Chat, such as ticketing, parking, working hours, and food and beverage options. It is likely because they can quickly get the information they need without navigating through the website or waiting on hold.”
Even more interesting, Arkharova says that customers might also feel more comfortable asking a larger number of questions when interacting with an AI. These interactions don’t carry some of the social considerations common to interpersonal interaction.
When speaking to an AI chatbot, a customer won’t be worried about taking up its time or asking certain questions.
But there’s a downside to this as well.
“Customers may be less likely to ask more complex or open-ended questions when interacting with AI Chat, as they may feel they need more confidence that it will provide a helpful answer. However, the technology improves so fast with the broader applicability of AI that the distinction between human and machine interactions will become less relevant.”
It’s another case of a current problem that’s already on its way to being solved, or at least alleviated.
The next step
That brings us to our closing question: what happens next?
“As more and more people start to use AI, the market for AI-based products and services will continue to grow, leading to increased revenue and market success for companies that can capitalize on this trend.”
We’re already seeing many companies take advantage of NLP capabilities, and even more brands will join the bandwagon in the near future.
Based on what we’ve seen so far, that’s good news for everyone involved, from customers to business owners to customer service reps.
“I think that AI chat models will continue to become more sophisticated and human-like, making it harder for customers to distinguish them from human representatives. This will lead to more natural and engaging conversations, resulting in improved customer satisfaction and increased sales.”
Conversational marketing represents a major step forward for the marketing landscape, and AI is chomping at the bit to move even further ahead.