There are more and more times when we find ourselves talking to what might seem like a person, but it’s actually a bot, often powered by AI. But what does the research say about the best way to use them? What can businesses do differently to improve their reputation, and get deeper customer insights? Professor Andrew Stephen, L’Oréal Professor of Marketing at Saïd Business School, takes us through some recent research his team has conducted around the impact of anthropomorphised (human-like) chatbots in customer service. Their assessment of thousands of real interactions, as well as some experimentation, has revealed that the more human a chatbot seems (even if the correspondent knows it is not human), the more we expect it to be effective. As he reveals, there are serious implications for brands in terms of how they deal with angry customers – not just in terms of resolving issues, but also related to longer-term brand preference and purchase intent. Julie Kollman, Chief Research Officer at Kantar, talks about some of the work we’ve been doing with chatbots in actually gathering insights. Replacing existing survey tools in the right occasions, chatbots have proven to be highly effective at getting us to deeper, richer insights – and probably more truthful responses. She explains some of the use cases, the cultural differences when it comes to willingness to engage with a research bot, and the way the technology is constantly advancing to provide ever better insights.