I really appreciate the conversation topics and reporting features in Intercom. However, these would be even more valuable if the machine learning algorithm also analyzed ticket attributes, not just conversation messages. In our workflow, users fill out detailed forms when submitting tickets, so a lot of important information is stored as ticket attributes. Often, if a customer’s report is clear, we do not start a separate conversation.
Currently, this means tickets are not automatically tagged with the appropriate conversation topics, even if the relevant keywords are present in the ticket attributes, since the machine learning only searches the conversations themselves. Expanding the analysis to include ticket attributes would ensure better topic tagging and more comprehensive reporting.