Real psychological safety with KPIs pairs honesty with accountability. Here are the eight qualities that make it possible, and where to start.
One of our PuMP alumni once said something I’d never heard put quite that way before: “PuMP is a psychologically safe way to develop metrics.” She meant it as a compliment, but at the same time it points to something we all know but rarely openly discuss: that performance measurement, done the usual way, often isn’t psychologically safe at all.
What psychological safety means for KPIs
The term comes from Harvard Business School professor Amy Edmondson, who defines psychological safety as the belief that you won’t be punished or humiliated for speaking up with ideas, questions, concerns, or mistakes. In the context of performance measurement, that means the belief that you won’t be punished or humiliated for the unacceptable results of any performance measure or KPI.
Psychological safety, in the context of KPIs or not, doesn’t mean taking a soft line in managing team performance. When Google ran its now-famous Project Aristotle study to work out what separated its best teams from the rest, they expected to find it in team composition or individual talent. Instead, psychological safety came out on top, ahead of four other factors that mattered too: dependability, structure and clarity, meaning, and impact. Teams with high psychological safety outperformed those without it.
Safety isn’t a soft option
Edmondson has spent a fair amount of her career correcting a misreading of her own work. Psychological safety, she says, isn’t about lowering standards, avoiding hard conversations, or wrapping a team in cotton wool. It’s the opposite of complacency. It’s what makes honest, rigorous conversations about performance possible in the first place, because people aren’t spending their energy on self-protection instead.
So psychological safety, especially in the context of measuring performance, isn’t about removing accountability for performance, but rather making it okay for people to share the truth that leads to real performance improvement.
That means psychological safety and accountability aren’t opposites. They work in harmony, not tension. A team can’t be honestly accountable for its KPIs if admitting a shortfall feels dangerous. (I’ve written before about what KPI accountability should actually mean, if you want to go deeper on that.)
Brook Rolter makes a similar point: when people feel unsafe telling the truth about performance, organisations lose visibility into how they’re really performing.
How fear creeps into performance measurement
Fear of KPI accountability creeps in through a fairly predictable path. Measurement will always start as monitoring to know what’s happening. But the moment a number becomes evidence for judging someone, monitoring turns into threat. And once people feel judged, they stop being honest, they get defensive, and the very data leaders need to see clearly becomes the data most likely to be hidden or reshaped.
It becomes a downward spiral where monitoring drives fear, which drives worsening performance, driving even more monitoring. If you’ve seen it happen in your own team and want to understand the mechanics of it, there’s more about this downward spiral here.
Mark Hocknell has written about this fear phenomenon directly, describing what he calls PM’O’Phobia — the fear of performance measurement.
What performance does when safety is missing
Without psychological safety, this downward spiral takes form at every level of every organisation:
- Employees game the numbers, the data, or the underlying process, because missing a target feels like an admission of failure rather than information to act on.
- Executives choose vanity metrics over honest ones, because sharing bad news puts their remuneration or their seat at the leadership table at risk.
- Governments publish only activity measures, not outcomes, because the scrutiny of media and community groups feels too exposing.
Every one of these is a rational response to feeling unsafe. That’s worth sitting with for a few moments. People aren’t behaving badly because they’re bad at their jobs. They’re behaving exactly as anyone would when the cost of honesty feels too high.
Eight qualities of psychologically safe KPIs
You may or may not have the power to change this in your organisation. But if you were to try, there are eight qualities your measurement approach will need for it to be psychologically safe:
- Focused on measuring business processes, not about measuring people — and this is a fundamental requirement until the organisation can guarantee the five conditions required support KPIs for personal performance.
- Structured to guide the steps required, not dependent on your current measurement know-how — and PuMP is such a structured method, giving teams the tools they need to create the structure & clarity Google identified as vital to team effectiveness.
- Structured to guide the thinking required, not limited by vague goals or legacy measures — this is where Google’s meaning dynamic shows up, where people find purpose making the right results better, not in a scoreboard.
- Played like a team sport, not resting all on your shoulders, alone — this echoes dependability from Google’s research, where team members work together to get things done on time and to a high standard.
- Emphasising feedback and learning, not about success or failure — the idea that is central to the entire psychological safety dynamic, which Google also ranked as the first dynamic of effective teams.
- Driven by continuous improvement, not about hitting targets — which touches directly on Google’s impact dynamic, giving teams to see how their work matters and how they have influence over improving it.
- Prioritising ownership and buy-in, not pushing technical sophistication too soon — since measurement doesn’t come naturally to many people, they can only become technically proficient if they first feel the relevance to measuring better.
- Allowing iterative progress, not expecting immediate mastery — because true performance improvement is often hampered by old practices that are tricky to unlearn before new practices can be embedded.
These qualities are all designed into how PuMP works. As long as leaders can model the same principles and practices that PuMP offers, every team can use it and feel safe in sharing the honest bad news that will lead to true performance improvement.
Ayça Tümer Arıkan has seen this shift firsthand — what changes when measurement stops just being a report and starts being a conversation.
Where to start
To create the conditions and benefits of psychological safety with KPIs, there are few places to start, regardless of where your organisation sits today:
- Change the question after a missed target. Instead of “why did you miss it,” try “what does this tell us, and what will we try next?” The first question invites self-preservation. The second invites thinking.
- Model fallibility before you ask for it. Edmondson’s own guidance to leaders: admit your own mistakes and uncertainty first. If a leader never says “we missed that target because we didn’t find and fix the real cause,” no one else will either.
- Open performance conversations with curiosity, not verdicts. A team performance review that starts with “what happened and what can we learn?” rather than “who’s responsible?” changes what people are willing to say in the room.
- Start with one of the above eight qualities that is missing from how your organisation measures performance. What would it take to design it back into your performance measurement and management approach? You can choose from 13 practical tips from how we build psychological safety with PuMP, here.
Kathy Letendre has documented what these alternatives can look like in practice — a team’s journey from burnout to breakthrough once measurement stopped being used against them.