Leadership & Behaviour

The Four Frictions: HBR's Test for Leaders Branded as Problems

A senior leader sitting across a glass-walled meeting room, mid-conversation with a sceptical colleague

Luis Velasquez, writing in Harvard Business Review on 7 May 2026, opened a question most leadership programmes quietly avoid. When a strong leader starts creating friction, the organisation almost always defaults to a single explanation, which is that the leader needs to change. Velasquez argues that the friction is usually not one thing. It is one of four, sometimes a combination, and getting the diagnosis wrong is how high-impact leaders quietly get pushed out.

The four sources he names are capability, perception, identity and system. The piece opens with Anna, a high-tech executive whose CEO admired her decisiveness and clarity, and who then began receiving complaints that she "moves too fast" and "makes decisions before the rest of us are ready". The CEO's instinct, predictably, was to put her on a development plan aimed at her behaviour. The actual problem, on Velasquez's reading, sat somewhere else.

Why this matters in plain terms: organisations are running succession decisions, promotion calls and performance plans on a single-cause model of leader friction. The model is wrong often enough that the cost shows up later as stalled pipelines and quiet attrition of the very leaders the business needs. Velasquez's contribution is to give those decisions a sharper diagnostic before they get made.

The Four Sources, In Practice

Velasquez's four labels are useful precisely because they look similar at the surface and resolve differently in the conversation.

The diagnostic discipline Velasquez is pushing for is to slow the conversation down long enough to ask which of the four is in play, rather than treating "the leader is the problem" as a single hypothesis with one tool attached.

Why The Default Is Always Capability

There is a structural reason most leader-friction cases get filed under capability, and it is uncomfortable. Capability is the only one of the four that can be addressed without the organisation looking at itself. Perception requires senior stakeholders to admit the read was wrong. Identity requires the unstated template to be made visible. System requires the leadership team to own a decision they would rather not own. Capability lets the friction be a one-person problem with a one-person fix. It is the cheapest diagnosis to make, and the most expensive to be wrong about.

The cost shows up in three predictable places. Promotion decisions get made on the wrong inputs. Succession pipelines stall because the high-impact leaders most likely to surface friction are the ones most likely to be filtered out. And the talent layer below learns, quickly, which kinds of behaviour the system rewards and which it punishes. Two of Sidestream's recent posts touch the same wound from different angles. The Edelman Trust Barometer 2026 finding that 42% of employees would rather switch departments than report to a manager with different values is, in Velasquez's language, a perception-and-identity case wearing a capability mask. DDI's Global Leadership Forecast 2026, with only 8% of executives rated strong at leading change, is what you get when capability development is asked to do the work of all four diagnoses at once.

What Most Organisations Try, And Why It Misses

The standard response to a leader creating friction is a 360, a coach, and a development plan. None of the three are wrong, all of them assume capability is the source. The 360 reflects perception back at the leader as if perception were the leader's own data to manage. The coach works on behaviours that, if the source is system or identity, were never the lever. The development plan creates a paper trail that, six months later, becomes the basis for the exit conversation.

What is missing is a forum in which the friction itself can be looked at honestly by the people creating it, the people receiving it, and the leader being labelled. That forum needs to be safe enough that the perception and identity sources can be named without anyone being the villain, and structured enough that the system source can be made visible without collapsing into politics. Almost no organisation has a venue for this conversation, which is part of why the default keeps being capability.

What Actually Works

Two findings from the wider behavioural-science literature are relevant. The first is Amy Edmondson's 1999 work in Administrative Science Quarterly on psychological safety, which establishes that whether a team will surface a misread depends on the senior person's first reaction, not on the team's intentions. The second is the practice gap. The conversation Velasquez wants leadership teams to have is not a conversation any of them have rehearsed. It is run, when it is run at all, under pressure, in the wrong format, with the wrong people in the room.

In Sidestream's own academic behaviour-change work, building on research from UCL, Cambridge and Bocconi, participants who learned a communication skill through immersive role-play scored roughly 20% higher on observed behaviour than those who learned the same content through video or slide-show training. Self-rated confidence did not predict observed performance: a Dunning-Kruger pattern we designed out of subsequent studies by replacing self-reports with behavioural measurement. The implication for the four frictions is direct. The diagnostic conversation has to be rehearsed before it is run live, with the same realism that a difficult board meeting would get.

What We Do About It

Our Change Resilience Lab builds a bespoke simulation around the leadership team that is actually running the friction. Professional actors play the stakeholders who are reading the leader as decisive, the ones reading her as overbearing, the colleague whose system change she is enforcing, and the leader herself. The leadership team has to diagnose, in real time, which of the four sources is producing the noise and what the right next move is. The room learns to slow the conversation down enough to make the diagnostic before the default kicks in. The same scenario gets run more than once. By the third pass, the four-source test is the new default.

When a leader gets branded a problem, the cheapest diagnosis is always capability. It is also, four times out of five, the wrong one.

The four-source test is not new psychology. It is a discipline of slowing down a conversation the organisation has already decided to have. Read our piece on the supporting-character leader for the same theme from a different angle, or book a call to look at what the diagnostic would look like in your team.

Book a free 30-minute diagnostic call →  or read about our research-backed approach.

Continue Reading

Related Articles

Leadership & Behaviour

The Supporting-Character Leader: HBR's New Test for Effective Managers

Leadership & Change

The 8% Problem: What DDI's 2026 Forecast Says About Change Leaders

Sidestream

Take Action

Bring Us Your
People Problem

Free 30-minute diagnostic call. No deck, no hard sell, just an honest conversation about whether we can help.