Balancing positive and constructive feedback using the Gottman ratio

Executive overview

Employee engagement has stagnated at 23–30% globally for over a decade, not because leaders lack data, but because management behaviour never changed. Well-being goes further than engagement: it asks whether people feel they belong, can be themselves, and are genuinely cared for — not just whether they're productive.

The core lever is a simple ratio. Research shows humans need at least four positive emotional experiences for every negative one just to sustain a relationship. Most managers invert this by default, fixating on what isn't working.

The fastest path to team well-being is a consistent 4:1 ratio of positive to corrective interactions, delivered separately — not sandwiched.

Why engagement initiatives keep failing

• Measuring engagement without changing management behaviour produces no improvement
• Surveys run once or twice a year deliver stale data, often months after the fact
• Without accountability, employees grow more cynical each survey cycle — worse than not asking
• Wall Street never treated engagement as a business metric, so CEOs and managers didn't either
• The system confuses measuring the problem with solving it

Engagement vs. well-being

• Engagement is what organisations want: discretionary effort, productivity
• Well-being is what employees need: growth, belonging, psychological safety, care
• Engagement surveys are inherently one-sided — they extract effort, not mutual investment
• Well-being asks: can I be myself here? Do my colleagues welcome and know me?
• 70% of engagement is driven by a person's direct manager, not organisational culture

The science of positive emotions

• Up to 95% of human behaviour is driven by feelings, not rational thought
• Barbara Fredrickson: humans are hardwired to thrive on positive emotions
• Any experience of positive emotion — joy, appreciation, interest, awe — registers as love
• The brain reacts to negative stimuli at least four times more strongly than to positive ones
• This negativity bias is why the ratio must be heavily weighted toward the positive side

The Gottman ratio applied to management

• John Gottman's research on long-term relationships: a 4:1 positive-to-negative ratio sustains them
• Below that threshold, even small negative interactions erode the relationship
• The same ratio applies to manager–employee interactions: drip positive emotions consistently
• Recognition doesn't cause people to coast — when it's sincere, people want to do more
• Poll surveys (frequent, short) give managers real-time data on whether they're hitting the ratio

Separating positive and corrective conversations

• The feedback sandwich (positive → negative → positive) backfires: people feel set up and manipulated
• Roy Baumeister's research shows people resent it — it reads as insincere
• Better approach: keep corrective conversations focused solely on the problem and the path forward
• Positive recognition should happen separately, regularly, and independent of any performance issue
• Frame corrective conversations around standards and solutions, not comparisons to peers

Building team cohesion over individual competition

• Comparing individuals to each other signals they are competitors — they stop sharing knowledge
• Holding everyone to a clear standard (not to each other) lets managers direct people to help one another
• Daniel Coyle's research on elite sports teams: no lone wolves; coaches who develop the whole person win consistently
• Focus on the team's collective performance; still acknowledge individual stars, but frame it within team success
• A cohesive team elevates individual performance more than ranking systems do

Making positivity genuine

• Leaders who ask "I don't want to be insincere" often don't intrinsically enjoy managing people — worth examining
• Genuine care shows up as small, consistent acts: a note, a forwarded article, noticing someone's effort
• One Academy member committed to one act of thoughtfulness per day; received responses saying "no one has said this in 20 years"
• The CEO example: "How are you really?" as a standard question signals real interest, not performance
• Positivity has no cost — the belief that recognition depletes something is a cognitive distortion

Starting points for leaders

• Audit your current ratio: assume you're lower than you think — most managers are
• Separate recognition from correction completely; never combine them in the same conversation
• Find seven things going well before raising the one that isn't
• Build team belonging actively: create conditions where people can be themselves without fear of diminishment
• Use short, frequent check-ins rather than annual surveys to recalibrate in real time