The IBM Chief Analytics Office (CAO) group is responsible for partnering with business units across IBM to embed AI and advanced analytics into workflows and deliver data-informed insights and machine learning recommendations, driving value for business areas ranging from ecosystem, product management, finance and more. For a global organization with over 250,000 employees, this is no small task.
Across IBM, there are more than 500 product managers supporting software product development. Collectively, they received over 30,000 new ideas and feature requests from users during the past year. To put these requests into action, they are categorized so that common themes can emerge and features can be prioritized. Historically, the feature request form contained an optional “theme” field that allowed users to self-identify a theme for their idea. But most ideas lacked an assigned theme, requiring product managers to manually review and categorize each feature request.
Within IBM, various product teams employed disparate approaches to evaluate new ideas and relied on separate platforms to prioritize them. This led to inconsistencies in selection criteria and data quality. Product managers stored configurations, data outputs and complex file formats in a local directory. Not only was this resource intensive, but it prevented file sharing between collaborators and often created accidental duplicates.