CASE STUDY · HR TECH / PEOPLE ANALYTICS PLATFORMS
Automating performance review workflows for a scaling company.
- 61%Review completion up from 61% to 78%
- 6 minutesAbout 6 minutes saved per review
01 · PROJECT OVERVIEW
The client and the brief.
An anonymised scaling company. Name withheld under confidentiality.
A review workflow with reminders and calibration. Employees and managers complete reviews in a single place. Automated reminders go out before deadlines. A calibration step lets managers and HR compare ratings across teams before they are finalised.
- 1Performance reviews ran on spreadsheets.
- 2HR sent forms by email, chased late replies by hand and merged results at the end of each cycle.
- 3Managers did not know what their peers were rating, so scores were uneven across teams.
- 4Reminders depended on one person remembering.
- 5Completion was low, and the cycle always overran.
02 · WHAT WENT WRONG
What went wrong, and how we fixed it.
Every engagement hits problems. These were the ones that mattered here, and what we changed so they did not come back.
- PROBLEM 01In the first live cycle, some managers released reviews to employees before calibration had finished, which caused confusion when ratings changed
We had assumed managers would wait for the HR signal. We added a hard lock: reviews cannot be released until calibration is marked complete by HR. We also added a clear status line on each review, so managers could see why release was blocked.
- PROBLEM 02The issue did not recur in the following cycle
- PROBLEM 03A second problem arose around reminders
Managers received so many that they began to ignore them, and some went to spam. We reduced them to three per stage, grouped by manager, and sent a weekly summary of who was outstanding. We also asked IT to allow the sending address.
- PROBLEM 04Completion started to rise after the change
03 · THE PROCESS
How we ran it.
We started by walking through the previous cycle with HR and several managers to see where time was lost. We agreed a simple flow: self review, manager review, calibration and release. We built it in weekly sprints and tested with one department before the full launch. HR ran a short training session for managers. During the live cycle we checked completion daily at first, and sent HR a summary of who was behind. After the cycle, we ran a retrospective with HR and managers and made changes for the next one.
- 01Step 1
We started by walking through the previous cycle with HR and several managers to see where time was lost.
- 02Step 2
We agreed a simple flow: self review, manager review, calibration and release.
- 03Step 3
We built it in weekly sprints and tested with one department before the full launch.
- 04Step 4
HR ran a short training session for managers.
- 05Step 5
During the live cycle we checked completion daily at first, and sent HR a summary of who was behind.
- 06Step 6
After the cycle, we ran a retrospective with HR and managers and made changes for the next one.
04 · RESULTS
What changed.
Measured against the way the work ran before the engagement began.
- NOW 01
HR no longer sends forms or merges spreadsheets.
- NOW 02
Reminders run automatically, and HR watches progress on a single screen.
- NOW 03
Managers compare ratings during calibration, so scores are more consistent across teams.
- NOW 04
The cycle ends on schedule and records are kept in one place.
- 61%
Review completion up from 61% to 78%
- 6 minutes
About 6 minutes saved per review
05 · WHAT WE BUILT
How it works.
- The application is a web app with single sign-on, backed by a relational database.
- Review templates are configurable, so HR can adjust questions and rating scales.
- A workflow engine manages states and deadlines, and a scheduler sends reminders by email and chat.
- Role-based access controls who can see a review: employees see their own, managers see their reports, HR sees all.
- Calibration views show rating distributions by team, with manager comments, and record every change with a reason.
- A bias check reports rating spreads by manager and by group, at an aggregate level, for HR to examine.
06 · QUICK ANSWERS
The short version.
- What problem did the client have?
- Performance reviews ran on spreadsheets. HR sent forms by email, chased late replies by hand and merged results at the end of each cycle. Managers did not know what their peers were rating, so scores were uneven across teams. Reminders depended on one person remembering. Completion was low, and the cycle always overran. As the company grew, the manual approach got worse each round, and managers complained about the time it took.
- What did Brandsmashers build?
- A review workflow with reminders and calibration. Employees and managers complete reviews in a single place. Automated reminders go out before deadlines. A calibration step lets managers and HR compare ratings across teams before they are finalised.
- How was the work run?
- We started by walking through the previous cycle with HR and several managers to see where time was lost. We agreed a simple flow: self review, manager review, calibration and release. We built it in weekly sprints and tested with one department before the full launch. HR ran a short training session for managers. During the live cycle we checked completion daily at first, and sent HR a summary of who was behind. After the cycle, we ran a retrospective with HR and managers and made changes for the next one.
- What were the results?
- Review completion up from 61% to 78%; about 6 minutes saved per review.
YOUR TURN
Have a similar requirement?
Tell us what is stuck. We will come back with a plan and matched, pre-vetted engineers.
