Organisational psychologist Tasha Eurich studied thousands of people and found a remarkable self-awareness gap: about 95% believed they were self-aware, but only an estimated 10–15% met her criteria — suggesting most of us understand ourselves, and how others see us, considerably less accurately than we think.

Most of us would probably say we know ourselves reasonably well. That confidence is exactly what makes organisational psychologist Tasha Eurich’s finding so uncomfortable.

In a research programme spanning ten investigations and nearly 5,000 participants, Eurich reported that about 95 per cent of people believed they were self-aware. Her team estimated that only 10 to 15 per cent of the people studied actually fitted its criteria.

The finding deserves attention, but it should not be read as a settled prevalence estimate for the whole population. The percentages came from a 2018 Harvard Business Review article in which Eurich explicitly called 10 to 15 per cent an estimate. That wording matters.

The figure came from a multi-part research programme

Eurich’s work was more substantial than asking a room of people whether they knew themselves. Her team reviewed nearly 800 scientific studies to examine the competing definitions of self-awareness. It surveyed thousands of people across countries and industries, collected ratings from people who knew them, and surveyed managers alongside their employees.

The researchers also developed what Eurich described as a seven-factor, multi-rater assessment because their review had not found a comprehensive existing measure that they considered sufficiently strong. A separate part of the programme involved detailed interviews with 50 people who appeared to have increased their self-awareness over time.

Those 50 had to clear four screening hurdles. They had to rate highly on the assessment, someone who knew them well had to agree, they had to report that their awareness had improved across their life, and the other person had to agree with that remembered pattern. This did not reduce self-awareness to private confidence.

There is still a limitation that often disappears when the headline number is repeated. Eurich’s 2018 article said the team was writing up its results for submission to an academic journal. I could not locate a later peer-reviewed paper publishing the full programme, sampling method and prevalence calculation. That does not make the estimate worthless, but it does mean 10 to 15 per cent should not be treated with the certainty of a precisely established population rate.

Knowing yourself and knowing your impact are different abilities

The most useful part of Eurich’s framework may be the distinction between internal and external self-awareness.

Internal self-awareness concerns how clearly people understand their values, passions, aspirations, fit with their environment, recurring reactions, strengths, weaknesses and impact on others. External self-awareness concerns how accurately they understand the way other people see those same things.

Eurich reported virtually no relationship between the two. A person can be highly reflective and articulate about what drives them while remaining inaccurate about how they come across. Someone else can read a room well and understand other people’s opinion of them, yet feel surprisingly unclear about their own values or direction.

Consider a manager who sees herself as decisive and direct. She may have examined her intentions carefully and know why she dislikes vague instructions. Her team may nevertheless experience the same behaviour as abrupt or dismissive. More detail about her intentions cannot answer the external question. For that, she needs information she does not control.

Eurich arranged the two dimensions into four broad archetypes. “Aware” people score highly on both. “Introspectors” have internal clarity but limited knowledge of how others see them. “Pleasers” understand their external image better than their own inner priorities. “Seekers” have room to develop on both dimensions. The labels are a framework, not permanent personality types.

The two percentages are measuring different things

The 95 per cent figure captures what people believe about themselves. The 10 to 15 per cent figure reflects Eurich’s estimate of who met her programme’s criteria. They are not two measurements made with the same instrument, and subtracting one from the other does not produce a scientifically precise “self-delusion rate”.

I think this is where the result becomes more useful, not less. Self-awareness is an unusually awkward quality to measure through self-report because the ability under examination is also the ability being asked to judge itself. Saying “I accurately understand myself” supplies confidence, but it cannot independently establish accuracy.

Workplace position complicates matters further. Eurich argued that power and experience can obstruct external awareness because senior people receive less candid feedback and may overgeneralise from what worked before. Her claim was not that every experienced leader loses perspective. It was that status can remove some of the corrective information that would expose a mistaken self-image.

A broader 2023 review in the Annual Review of Organizational Psychology and Organizational Behavior reaches a compatible measurement point. Self-awareness includes how people see themselves, the effect they have on their environment and the accuracy of those perceptions. Reflection, feedback and interpersonal learning all matter because no single private viewpoint captures the full construct.

More introspection does not automatically improve accuracy

Eurich’s programme also challenged a comforting assumption: that enough thinking about ourselves will eventually make our self-image accurate.

In an HBR IdeaCast interview, she described finding that people who spent more time self-reflecting were not necessarily more self-aware. The issue, in her interpretation, was not reflection itself but the way it was conducted. Repeatedly asking why something happened can invite an explanation built from incomplete evidence, then make that explanation feel truer through repetition.

Questions centred on what happened, what pattern can be observed and what could be done differently keep attention closer to information that can be checked. This is a tendency reported within Eurich’s research, not a universal rule that one question word produces insight and another prevents it.

Internal reflection without an external check can become a very polished story. That story might even be sincere. Sincerity and accuracy are not always the same thing.

The finding is a reason for humility, not labelling

The tempting social use of the 10 to 15 per cent statistic is to sort other people. A difficult colleague becomes “unself-aware”, while the person making the judgement quietly places themselves among the perceptive minority. The statistic then recreates the confidence gap it was supposed to expose.

I recently wrote about selective attention as a mature form of self-awareness. Eurich’s distinction adds a useful complication to that idea. I can know why I no longer spend energy on every opinion and still be wrong about whether people around me experience that selectivity as calm judgement, distance or indifference.

The exact percentage will remain debatable until the underlying programme is available in enough detail to evaluate fully. The gap it points towards is easier to recognise: people have privileged access to their intentions, while everyone else has direct access to their behaviour.

A serious account of ourselves has to leave room for both.