Accelerating real time feedback at PwC with AI
Nudge employees to request and give feedback on a regular and timely basis
Overview
PwC sought to transform employee feedback from a "check the box" exercise into an ongoing, meaningful process. With thousands of employees and demanding workloads, the company needed a way to keep feedback top of mind in employees.
Goal
Increase Snapshot initiation by PwC employees. Change the behaviors and perceptions around performance evaluation from "high stakes, administrative burden" to frictionless and natural.
Lead UX designer - drove the project and product strategy, supported research
1 PM, 2 UX reseachers, 2 UX designers, 1 UI designer
Figma, Chat GPT, Miro
Sept 2024 - Dec 2024
The Problem
Employees weren’t consistently requesting or completing Snapshots (PwC's main feedback and performance mechanism), missing valuable feedback opportunities. Key challenges included:
Delayed Feedback: No system encouraged real-time insights, impacting career growth and team dynamics.
Competing Priorities: Heavy workloads pushed feedback requests to the back-burner.
Manager Overload: Sporadic requests caused an influx at performance review periods, overwhelming leaders.
To make feedback natural and low-friction, we needed a solution that fit seamlessly into employees' workflows.
Scope and Constraints
Scope: Increase Snapshot initiation by integrating it into the Astro platform.
Constraints:
Designed as a streamlined add-on, not a replacement for the larger Snapshot Tool.
Needed to integrate smoothly into Astro without disrupting core workflows.
AI-driven suggestions required balancing automation with user control while minimizing biases.
Had to be frictionless, avoiding added complexity or administrative burdens.
Key Features
Seamless Integration
Most PwC employees are in Astro daily to complete their time entry requirements. We incorporated feedback reminders with the time entry success confirmation. You logged your hours; now get credit for the work you did. We established logic of when the banner should appear to ensure relevance and avoid banner blindness. We used imagery synonymous with AI so that employees knew they wouldn't be writing from scratch.
Banner appears in Time Entry confirmation screen with an entry point into the Snapshot Assistant
Gen AI Assistance - the "Improve It" button
Many users cited the “blank page” as one of the most difficult parts of completing a Snapshot. They also were frustrated by the low character limit of Snapshot descriptions. Our implementation of Gen AI allowed them to write as messily as they wanted and be as verbose as they needed. Afterwards, the implemented prompt would tie their work back to the PwC professional model and make the description as concise as possible.
The user's improved Snapshot Description appears. They can replace it or retry.
Encouraging Face to Face Interactions
We used our previously successful “schedule time” feature to encourage employees and reviewers to discuss the performance and feedback. Once an employee selected a reviewer, Astro looked for an open meeting time for both and prepared the calendar invite. The implementation of a default scheduled meeting meant users would have to opt out if they didn’t want to book a one-on-one.
Automatic booking - first available time selected with 2 alternatives offered.
THE RESULT?
1,200 new Snapshot requests within the first two weeks of the Snapshot Launch at the beginning of the New Year.
This was significant as January marks the beginning of the second half of PwC’s performance year and is typically a low point for feedback requests.
Key Takeaways
Habit-stacking Increased Adoption: Marrying Time entry + Snapshot changed perceptions around feedback frequency.
Balance Automation & Control: Users valued AI-assisted feedback but wanted the final say.
Leverage Familiar Workflows: Embedding Snapshots into an existing tool and using familiar patterns reduced friction and lowered the perceived effort of feedback requests.
Next steps
After this successful collaboration between the Astro team and Career Development team, we plan on exploring how else we can integrate feedback processes into the Astro platform and where else we can leverage AI throughout the career development experience.
