MARA ULLOA HCI Researcher

/ MAHR-ah - OOh - YO-ah /

Product Managers’ Integration of Generative AI into Knowledge Work

April 2025 - June 2025 (12 weeks)

Project overview

Microsoft leadership wanted to investigate the use of Generative AI (genAI) by product managers (PMs), a traditionally understudied user group. Previously, only software engineers had been involved in the conversations about genAI usage at the organization. As this was the first comprehensive study on PMs and genAI, I designed and led a mixed methods study, distributing a large scale, global survey (885 participants), in parallel with semi-structured interviews (15 participants) and triangulated with telemetry data.

Process and Methods

  • Internally validated research instruments with user group and other productivity researchers
  • Led data collection
  • Collaborated across disciplines (data science, product management, software engineering, and AI research)
  • Qualitatively and quantitatively analyzed data across sources
  • Developed Python charts to visualize quantitative and qualitative analytical findings
  • Performed thematic analysis on interview and open-text survey questions
  • Conducted workshops with product managers to refine findings

Deliverables

  • Presented findings and their implications to leadership (corporate vice presidents, managers, AI + marketing)
  • Produced the Selective Delegation framework to depict how PMs delegate tasks to genAI
  • Reported on PM GenAI adoption rates, use cases, and perceived benefits and challenges
  • Guiding the production of a playbook to augment PM genAI literacy at the company
  • This work forms part of Microsoft's New Future of Work Report
  • First-authored the below academic paper to disseminate findings and implications of this research

Research Paper

Product Manager Practices for Delegating Work to Generative AI: ‘Accountability must not be delegated to non-human actors’

🎉 this paper has been accepted to ACM/IEEE International Conference in Software Engineering 2026

I conducted informal interviews, 15 interviews, and gathered 885 survey responses. 731/885 survey participants consented to system telemetry data analysis.

The study design

PMs identified values at the individual, team, and organizational levels influencing task delegation to genAI

The Selective Delegation Framework

Survey analysis from 885 PMs, including both Indvidual Contributors and People Managers

Survey analysis of 885 PMs on genAI job displacement cocerns and genAI usage encouragement

Individual Contributor coded responses to the optional,open-text question: Is there anything else you’d like to tell us about barriers to using GenAI more effectively in your role?

Individual Contributor coded responses to the open-text question: Is there anything else you would like to tell us about barriers to using GenAI more effectively in your role?

Individual Contributor coded responses to the optional,open-text question: In your opinion, what is the single biggest benefit you’ve gained from using Generative AI, if any?

Individual Contributor coded responses to the optional,open-text question: In your opinion, what is the single biggest benefit you've gained from using Generative AI (if any)?

A selection of responses to Likert questions posed to 735 Individual Contributors about their thoughts and beliefs on GenAI.

Individual Contributor coded responses to the optional,open-text question: In your opinion, what is the single biggest benefit you've gained from using Generative AI (if any)?

Selected Projects