I have recently come across an interesting paper examining under what conditions customers preferred to interact with a human vs a non-human agent, after a failure with self-service technology (SST). That is, if a customer is using technology to perform a task (say, ordering something online), and something goes wrong, would customers prefer to be handed-over to a member of staff to fix the problem, or have the problem be solved by the/a non-human agent?
As someone who has had some frustrating interactions with self-service technology recently, I was very curious to know what the authors had found.
The paper I am referring to is entitled “Self-service technology recovery: the importance of psychological need support”. It is authored by Byron W. Keating and Marjan Aslan, and it was published in volume 34, issue 4 of the Journal of Service Management.
The authors conducted a series of online experiments, documented in the paper, and found that customers want a SST service failure to be resolved by technology. Obviously, they prefer to be helped by a human agent than receiving no help at all; and, naturally, the preference for human vs non-human agent is impacted by personal and task characteristics (e.g., how complex the task is). But, the bottom line is that, once a customer is down the path of using SST, then firms had better ensure that their technology works and that can help customers undo whatever problem the SST has created.
The authors posit that this is because “users experience a service obstacle as thwarting their locus of causality, robbing the user of the task-inherent feedback that would result from being able to complete the desired service task independently. Any assistance provided, therefore, needs to redress this loss and restore a user’s locus of causality” and recommend that “service providers could employ non-human technologies such as AI-enabled chatbots to provide automated feedback addressing common problems to allow users to self- manage their recovery. This strategy would benefit from reinforcing perceived competence and is likely to fit well with a user’s preferred interactional style (i.e. technology-based self- service).”
An important caveat of this study is that it looks at volational use of SST, only – i.e., when the customer wanted to use the SST, in the first place (as opposed to being forced into using it). Still, I found this research finding interesting.
Did you have interesting SST service recovery situations, lately?
Ana Canhoto 
If you put the question differently, you maybe get a different answer. So, if you were given the choice between a chatbot to resolve your problem and a real person on the phone, I would always go for the former. I have had a fear-hate relationship with the phone all my life and have now found it is a condition, telephobia. I would do almost anything rather than pick up the phone. So back to your question, Ana: I would happily deal with a human online, in a chat where I can discuss things, give the details and then save a transcript of the chat.
LikeLike
You are “in with the kids”, Robin – the young adults who seem to prefer interacting with each other via text than voice 😉 Joke aside, you raise a very valid point: i.e., tech as a medium vs tech as an agent. I suppose such preference could fall under the boundary conditions of “consumer characteristics”, but I don’t think that the authors have considered it, specifically.
LikeLike