A study comparing human–human interactions to human–chatbot interactions, found that the latter typically last longer than human–human interactions between strangers. They also involve shorter messages, less complicated vocabulary, and even more profanity! 1
There is emerging thinking that chatbots can be harnessed by the research industry. They present an opportunity for quick on the spot data. But just how useful are they if bots elicit a behaviour change as outlined above?
Many brands are starting to use chatbots for straightforward customer transactions with genuine success. In these transactions, customers just want to get the job done and whether it is by a human or a machine doesn’t matter. It makes sense that their implementation could be extended into simple research. This has a number of benefits.
1. The ‘conversation’ is instantaneous and in real time as it is from a mobile platform. The interactive nature of the ‘conversation’ mimics the spontaneity of the moment and arguably the feedback is more ‘system 1’ as the response happens off the cuff. It’s speedy and in the here and now.
2. It starts in consumers shoes and embraces the trend for messaging. This is a substantial benefit given the move from email as well as from social media. Peter Rojas of Betaworks states: “People are now spending more time in messaging apps than in social media and that is a huge turning point. Messaging apps are the platforms of the future and bots will be how their users access all sorts of services.” In the research scenario, it moves from being a mobile survey as such, to more of a mobile interaction.
3. The bot approach provides an opportunity in particular to connect with millennials who are typically not picked up by survey panels. They can answer with their favourite messaging app or service. In truth it reflects how the bulk of their personal remote conversations happen.
4. It avoids consumers having to download an app for a particular survey platform. This furthers the impression people are dealing with a mobile interaction rather than a mobile survey.
Nonetheless the current immaturity of this technology has to be borne in mind. At the moment its best use is limited to closed questions, yes and no answers and multiple choice.
Undoubtedly the user experience is also critical to get right. At face value the bot appears incredibly simple but there are many challenges such as flow optimisation and being compatible with ever-changing platforms. If people are finding themselves having to correct the ‘bot interaction’ and ‘fix’ misunderstandings, the spontaneity of the moment is affected and thereby realism of response.
We should also not forget consumer perceptions of the chatbot. From early adopters to the technical reticent, there are varying levels of comfort.
Indeed, the perceived novelty may mean the bot respondent uses and interacts with them in a different way. If the technology were to venture into the qualitative arena and open-ended questions, the response to a bot question may be different. The typically shorter interaction means less richness in the respondent linguistic variety which typically helps reveal hidden emotional drivers and barriers.
There is insufficient evidence currently to assess the traits of respondent behaviour via a bot-enabled survey given the nascent use of this technology. But it will be interesting to see how the research industry can embrace bot technology and whether it truly finds a place and adds value in the long term.
If you’d like to discuss further or explore any other insight methodology that might be appropriate for your organisation, please do get in touch for a no-obligation discussion firstname.lastname@example.org
1 [Hill, J., Ford, W.R., Farreras, I.G.: Real Conversations with Artificial Intelligence: A Comparison Between Human–Human Online Conversations and Human–Chatbot Conversations).