Two days before Mental Health Awareness Week opens on 11 May 2026, the headline number worth carrying is one most boards have not been quoted. ACAS's Prevalence of Conflict at Work report, published 20 November 2025 from a sample of 4,558 working-age adults in Great Britain, found that 44% had experienced conflict at work in the previous twelve months. ACAS itself describes this as "the highest level of individual conflict at work ever reported in a survey of workers in Great Britain." Earlier estimates, using different methods, sat between 25% and 38%.
The headline cost figure ACAS cites in the same report is £28.5 billion a year in lost output across Great Britain. That is not a wellbeing line item. It is a P&L one.
What The ACAS Survey Actually Found
The 44% prevalence is a single number on the cover. The detail underneath is more useful for anyone designing the response.
The five most common topics, in order, were capability and performance (38%), personal disagreements and relationship breakdowns (33%), bullying, discrimination and harassment (24%), employment terms and conditions (17%), and misconduct (14%). The top two, between them, account for the majority of cases, and both are about how managers and colleagues talk to each other under pressure rather than about policy or procedure.
The second number worth pausing on is the impact one. 57% of those who experienced conflict reported stress, anxiety or depression as a consequence. That is the link Mental Health Awareness Week, with its 2026 theme of Action, is asking employers to take seriously this month. The conflict and the wellbeing crisis are not parallel problems competing for budget. The first is producing a meaningful share of the second.
The third is harder to act on but harder to ignore: nine in ten potential tribunal claims are still resolved through ACAS early conciliation, and from December 2025 the early conciliation window doubled from six to twelve weeks. The legal off-ramps are working. The behavioural ones, the conversation that prevents the dispute, are the ones that have not been built.
Why The Standard Diagnosis Misses
The default response to a record conflict number is procedural. Update the grievance policy. Re-launch the dignity-at-work training module. Add a respect statement to the values deck. None of those touch the moment a manager realises a personal disagreement on their team has stopped being about the work.
Two findings from the wider research literature explain why awareness materials so rarely shift conflict numbers. The first is Roediger and Karpicke (2006), in Psychological Science: being asked to retrieve and apply material increases long-term retention by around 50% compared with re-reading the same content. A two-hour module on having difficult conversations behaves like the re-reading condition. Eight weeks later, the manager who is sitting between two team members who have stopped speaking is operating on instinct.
The second is Edmondson (1999), in Administrative Science Quarterly, on psychological safety, defined as "a shared belief that the team is safe for interpersonal risk-taking." The behaviour ACAS is measuring, the disagreement that escalates rather than resolves, lives almost entirely in the gap between what people think and what they feel safe saying out loud. Google's Project Aristotle, the multi-year study of team effectiveness, ranked psychological safety first among the dynamics that distinguished higher-performing teams. It is the same construct, reached through a different door.
What Works
The organisations whose conflict, grievance and absence numbers have moved downward share three habits.
They define the behaviour, not the policy. The point is not whether managers know what bullying is. The point is whether they can, in the actual meeting, name the behaviour they have just witnessed, ask the question that interrupts it, and stay in the room. Those are skills, and skills decay without rehearsal.
They put the difficult conversation in front of managers in a low-stakes setting before the real one arrives. Building on academic behaviour-change work from UCL, Cambridge and Bocconi, our own research found that immersive role-play with professional actors was around 20% more effective than passive modalities such as slide-show or video e-learning at moving observed skill. The same study found participants overestimated their skill before measurement, and underestimated it after, the Dunning-Kruger pattern that quiet conflict thrives on.
They re-measure. Not training delivery, not awareness scores. The honest test is whether grievances raised in month nine are smaller, earlier and resolved before they reach the formal procedure. The £28.5 billion figure does not move because the conflict prevention conversation is the part of the year that always slips.
This is the work behind our difficult-conversation labs and our wider immersive simulations. Professional actors stage the moment, with all the social pressure intact, before the real version arrives unannounced on a Tuesday afternoon.
The honest test for any organisation reading the ACAS numbers this Mental Health Awareness Week is simple. Pick the most likely live conflict in your business right now. Ask the line manager involved when they last practised the conversation that would resolve it, in conditions close to the real ones, with feedback. If the answer is "never", the £28.5 billion is not a national figure. It is a local forecast.
Book a free 30-minute consultation → or read about our research-backed approach.