Not all consumption settings are the same. They can range from situations where we can easily evaluate what we are consuming (for instance, buying a new car), to situations where we can only evaluate what we are consuming afterwards (for instance, renting a car), as well as situations where we will struggle to evaluate what we have consumed even after we have done so (e.g., a car MOT). We say that the first type of setting is high in search attributes, the second one in experience attributes, and the third one in credence attributes. This difficulty in evaluating what we are getting for our money, makes the settings high in experience attributes riskier than those high in search attributes; and makes the third type of setting – the one high in credence attributes – the riskiest of them all. [see this previous post about the understanding risk, and its impact on consumption].
The paper “The adoption of AI service robots: A comparison between credence and experience service settings”, argues that the drivers of adoption of AI for experience settings are different from those for credence settings. The authors of the paper are Sungjun S. Park, ChunTing D. Tung and Heejung Lee. They conducted a study in South Korea whereby they randomly assigned participants to one of two scenarios: the first one, described an interaction with a robot in a café (which is an experience setting); the second scenario described an interaction with a robot in a hospital (i.e., a credence setting).
The researchers, then, asked various questions about the attributes of the technology, I including how easy it was to interact with the robot (i.e., its Ease of Use), and the extent to which it helped with achieving the desired outcome (i.e., its Perceived Usefulness).
They found that Perceived Usefulness was important in both settings, but more so in the Hospital setting. That is, the more research participants believed that using the robot would help them achieve their goal successfully, the more they were willing to use it. In this credence setting, Perceived Usefulness had a strong direct effect on behaviour intentions, as well as a very strong effect in shaping attitude towards Robot usage (which, in turn, had a determinant effect in subsequent behavioural intention). While Perceived Ease of Use increased the Perceived Usefulness, it had a negligible effect on attitude.
In turn, Perceived Ease of Use emerged as a key driver of robot acceptance in the café setting. That is, the more intuitive and seamless the experience of interacting with the robot was, and the lower the effort require to do so, the higher the willingness to use that service option. In this experience setting, Perceived Ease of Use continued to increase the Perceived Usefulness, but, more interestingly, also had a very strong effect in terms of shaping attitude towards Robot usage (which, again, had a determinant effect in subsequent behavioural intention).
In a way, these findings are quite intuitive: in a risky environment, we may be more concerned with how effective the robot is than how easy to use it may be. Nonetheless, these findings are still interesting in terms of showing the direct and indirect effects of these two perceptions on other variables which go on to shape behaviour.