Are interactive bots really the answer to every marketer’s prayers, or still a long way from being much more than a robotic nuisance? Anna Haverinen offers a design anthropologist’s view on the topic.

These days, conversational bots are all over the tech news. Silicon Valley has declared them as the next hot innovation of the future, while anthropologists have remained sceptic, but curious about their potential.

Artificial intelligence is deeply embedded in the technologies we are used to spend time with every single day. The bots are already not only ordering our taxis and delivering our pizzas, but also having racist meltdowns on social media and continuing our loved ones’ Twitter presence after they have passed away.

This begs the question: where does one draw the line between helpful and annoying, or innovative and downright creepy?

Poncho the weather cat
Lately, Facebook’s messenger has introduced a number of chat bots. Poncho the Weather Cat is one of our personal favourites.

When designing a conversational bot, there are several things to keep in mind. For starters, scripted bots are prone to serious “lost in translation” moments, if their dialog tree has not been tested enough. This may cause the customer to walk away, which makes it crucial to clearly communicate what the bot is capable of so that the exchange does not result in frustration. Moreover, interface design plays a significant part in how bot messages are interpreted by the users. Same content can end up conveying a different kind of message depending on the context and psychological effects of the UI.

Most bots are a long way from passing the Turing test, but they don’t necessarily have to. They simply need to make the user’s life easier.

Some claim that a well designed website or an app doesn’t need a chat bot. And often that really is the case. However, in many cases conversational bots can make the experience more smooth, lively and even personal – given that discussing with the bot feels smooth and delightful. Most chat bots are a long way from passing the Turing test, but they don’t necessarily have to. They simply need to exist for a reason: to make the user’s life easier.

Siri vastaa
Apple’s Siri does not care for your science fiction fantasies.

As an anthropologist I feel it’s also necessary to think about cultural contexts when designing the chatbots. There are differences around the world how people are accustomed to use apps and text-based services, especially in case if they try to mimic a person. This is why it’s a smart idea to come forth with the fact that the bot is actually a bot, and not a flexible human being. This way people don’t get as frustrated or annoyed, if they realise the bot cannot help them with a very specific need.

People also have different expectations when it comes to mobile and website bots. That’s why designers need to pay a lot of attention to the rhythm and timing of bot interventions. Nobody likes a histrionic bot that continuously fills your screen while you just can’t seem to find the X button that makes the damn thing go away.

Even with all their possible issues and quirks, bots can do awesome things and make our experiences with digital services more interactive. And perhaps the most human quality of the bots is the fact that they can be only as good as we make them.

Interested in real-life cases where we’ve dealt with interactive bots? As it happens, we’re currently helping our client Tidy Technologies to create a bot which helps website builders and editors to reach their site’s full potential. Read all about it here.

Suggested posts: