Chatbots are being heralded as the next big marketing development and it is only a matter of time before companies fully jump onto the bandwagon and get themselves a chatbot. Beerud Sheth, CEO of messaging platform Gupshup, has been quoted saying that “I think you’re going to see a bot explosion. I think it’s safe to say that this is the year of the bot. All the technology pieces are in place.”
A “chatbot” is a program, powered by artificial intelligence, which mimics conversations with people. Chatbots typically come in two forms: rule based chatbots and automated machine learning chatbots. A rule based chatbot takes a user’s written instructions and cross references them against its database of historic conversations, trying to match the instructions with previous answers. The chatbot then provides the user with what it believes is the most appropriate response. Machine learning chatbots try to learn from user conversations, in real time, to become smarter.
Chatbots rely on data volume to learn, make mistakes and then avoid repeating those mistakes. The original role of a chatbot was to provide simple customer support or gather customer feedback. Many customers will be familiar with the frustration that comes with calling a retailer’s helpline, only to find that you need to navigate a combination of options presented by an automated voice response. Chatbots were brought in to simplify this process and interact with users through various questions and answers, with the aim of more efficiently and quickly directing users to their requested information.
However chatbots can also offer retailers the opportunity to boost brand engagement by helping shoppers make product choices, check store locations and opening times, place orders or engage with campaigns. Chatbots give retailers access to a large user base and the chance to personalise each user’s experience, as well as engaging with a younger generation of users. Satya Nadella, the CEO of Microsoft, said at this year’s Build Developer conference that chatbots will sit alongside digital assistants and humans, forming a trio underpinning the future of computing. Digital agents, like Cortana, Alexa and Siri, will get to know users across a range of devices and understand their preferences, whilst chatbots will be used as the go-between the trio, customers and other systems. The trio can work together to plan a purchase or make a recommendation.
The chatbot revolution is no longer in its infancy; however it still has quite a way to go to remove the common issues associated with bots. The big players, notably Microsoft and Facebook, are making great strides in conquering the chatbot world. Microsoft is allowing users to converse with chatbots on Skype, with the aim of having Cortana (Microsoft’s digital assistant) as the interface between users and chatbots in future. Users will be able to tell Cortana what they need e.g. buy a book, and Cortana will pass on the relevant instructions to third party bots, such as book retailer bots. As a chatbot platform, Skype offers the possibility of users providing video or audio, rather than just text, instructions. Facebook has also launched its “Bots for Messenger” tool, which allows developers and businesses to build chatbots for Facebook’s Messenger platform. Facebook’s Messenger platform has over 900 million monthly active users and so has access to a huge user base. There’s already even a chatbot equivalent of an app store, courtesy of botlist.
However retailers should remain cautious when entrusting chatbots with their customer relationships.
- Adding value. Chatbots should add value to customers’ lives and not increase the levels of complexity, preventing customers from getting to the right answers. A common issue currently facing most chatbots is their inability to understand user instructions. The chatbots should be able to understand customers’ queries and provide helpful and appropriate suggestions. In addition, chatbots should learn how often customers want and need a response.
- Rogue chatbots. Retailers should only expose customers to their chatbots once they are confident that the chatbots will enhance, rather erode, their brand name and value. A chatbot’s personality should be reflective of the retailers’ brand; for example, a creative retailer, such innocent, is more likely to have a chatbot that’s quirky and humorous, as compared to a chatbot assisting a bank. Retailers should view their chatbots as an extension of their branding and tailor their personalities accordingly.
- Data is key – a chatbot is only as good as its underlying data. A significant advantage of chatbots is their ability to sift through a vast amount of data, so providing customers with comprehensive and up to date information. Companies will need to make sure that their chatbots are using current and complete data, such as inventory levels and product locations, when interacting with users.
- Legal issues – as chatbots become involved in more complex transactions and requests, a chatbot will need to be programmed to understand these instructions and progress with the correct outcome. Every potential outcome will need to have been considered by the chatbot’s programmer so that every eventuality is carefully coded into the bot. The chatbot will need to recognise when an instruction may require the customer to accept terms and conditions, provide these to the user at suitable time and keep a record of the terms accepted by the user. The chatbot may also need to share any information provided by the customer with third parties, and deal with the associated data protection issues accordingly.
- Moral issues – chatbots may be used to provide advice to users, for example through medical or financial advisor bots. Developers will need to provide their bots with sufficient data to enable them to work through the issues which users will want to discuss, and will need to be comfortable that their bots are able to properly handle these complex issues. If bots are advising on users’ health queries, life situations or financial decisions, the advice will need to be correct, responsible and highlight any associated risks.
Next time, we’ll consider the key legal issues faced by companies and chatbot platforms when advertising to consumers using this new channel of communication.