Don’t Ban the Bots: How to Make AI and Live Commerce Work Together

Artificial intelligence-powered chatbots may be extremely common in customer service, but do they make sense for sales? Consumers seem mixed on the idea, but overall, they seem to prefer bots for uncomplicated tasks. Salesforce, for example, found that 69 percent of consumers prefer to use chatbots for getting answers to simple questions. But when people are considering buying a product, they have a wide range of questions that are hard to anticipate and not always easy to resolve.

One unexplored answer to this dilemma is to start with live commerce — i.e., connecting customers to human agents via video. Live commerce can potentially be used to increase sales in the short term and build scalable AI in the long term.

To understand how this can work, let’s start with the reason we use AI chatbots in the first place: efficient scaling. As the volume of customers and queries increases, chatbots resolve a lot of easy questions without human intervention.

But while AI scales infinitely, live commerce also scales fairly easily. Unlike customer service, which often requires technical expertise, live commerce can not only take place in a scalable call center, but it can also be done by anyone in your organization using a distributed app. The majority of people may not be salespeople most of the time, but they can certainly jump in, as necessary. Scaling is, thus, a built-in feature of most brands.

We should also keep our eyes on the prize with scaling. The goal is not to increase the number of interactions, but the number of conversions. AI chatbots can scale infinitely, but that doesn’t matter if only a tiny percentage of them manage to complete a sale. Live commerce agents are proven to increase sales anywhere from three times to five times, and even relatively ineffective ones are more effective than nothing at all.

AI vs. live commerce is also not an either-or proposition. Rather, they can and should be complementary. Every interaction can be converted into text and image data that can aid in understanding the most common queries and sticking points for customers.

This data can add value in two main ways. The first is that whenever you launch a product, you may know what will entice people to buy, but you never know what will keep them from buying. You may have the perfect product that everyone needs, but if it’s too purple, that could be a problem. Live commerce is a great way to understand what people really want and what really concerns them — and to adjust sales training and product development accordingly.

The second is that live commerce produces an ideal training data set for AI because it uncovers real issues and the considerations that resonate with the shoppers who have them. This allows you to train your chatbots to handle a greater number of complex queries.

While we will never develop AI to be able to answer every question, we can develop it enough to increase scale and efficiency. To do so, we have to start gathering the live, granular, personal and intimate data we need to train them. Live commerce is, thus, an ideal place to increase sales in the short term and strengthen AI in the long term.

As a result, your AI sales efforts should always go hand-in-hand with live commerce. By analyzing tons of highly scalable conversations with real clients, you can uncover the blockers that can make AI more efficient and shift your live agents to more difficult and higher value conversations.

Dragorad Knezi is the CEO and co-founder of eyezon, a provider of live shopping experiences on-demand.