You are Priya Anand, lead of Atlas, a nine-engineer squad at Lumera Cloud, a Sydney software company. Your squad has missed three sprints in a row and has stopped giving one another honest feedback. By Friday your VP of Engineering expects your diagnosis: why did it stop?
The situation
Lumera Cloud has 410 staff, 240 of them engineers spread across 22 squads, and about A$96 million of annual recurring revenue. After its Series C round in 2024 the company rolled out PulseHR, a platform that scores every engineer weekly on delivery velocity, code-review turnaround and peer recognition, and feeds quarterly pay calibration. It was sold as a way to make performance conversations continuous and objective. The board paid A$1.1 million for it and has no intention of switching it off.
Atlas's numbers on 3 June 2026 are hard reading. Before PulseHR, each engineer asked a peer or a lead "how am I doing on this?" about 14 times a quarter; now it is 3. Trust in how performance is evaluated sits at 34 out of 100. Work marked done and later reopened has gone from 9% to 22%, and the time a sprint takes now varies twice as much as it used to.
Ten weeks remain in the delivery window. A fourth missed sprint would trigger a A$540,000 service credit with two customers worth A$7.2 million a year combined, and replacing a senior engineer costs about A$180,000. At this morning's retro, two senior engineers said, almost word for word, that nobody gives real feedback anymore because it all ends up in a score that sets pay. Others in the building read it differently: that PulseHR's dashboards measure the wrong things and are buggy, that the squad lacks places to talk, or that people simply stopped trying.