In order to help adolescents cope with loneliness during the social distancing and isolation imposed by the COVID-19 pandemic, we designed a conversational agent programmed to distract users from negative thoughts and advise them on strategies to improve their wellbeing. In order to assess the effects of the agent intervention on adolescent participants, we performed quantitative analysis of their self-reported mood states and qualitative analysis of their subjective views and opinions on the agent to help us understand their experiences. Trends in the quantitative data point to minimal changes in participants' wellbeing and loneliness after interactions with the experimental agent. However, qualitative data on adolescent experiences suggests short and long-term positive effects of the experimental interactions. In reporting our findings, we aim to bring attention to the importance of the qualitative data for understanding human experiences with technology, as well as the limitations of the instruments developed in the field of psychology for human-information interaction research.