The discomfort of looking at ourselves honestly

Let's start with the obvious: nobody enjoys finding out they are worse than they thought. A team's self-esteem, like an individual's, tends to protect itself. And there is a well-studied psychological phenomenon that makes the problem worse. David Dunning and Justin Kruger described in 1999 that people with lower competence in an area tend to overestimate their level, partly because they lack the very competence needed to recognize their shortcomings. The less we know about something, the worse we judge how much we don't know.

Applied to agility, the effect is uncomfortable. A team that has barely scratched the surface of what it means to work in an agile way can feel fully competent, precisely because it doesn't know how deep what it's missing goes. It has learned the visible mechanics, the ceremonies, and the vocabulary, and it confuses that mechanics with substance. Confidence, in this case, is not a sign of mastery but sometimes the opposite.

What we say we do versus what we do

Chris Argyris devoted much of his work to a distinction that sheds light on this problem. There is a systematic difference between our espoused theories, what we say we do and believe we do, and our theories-in-use, what actually governs our behavior. They almost never match, and the most striking thing is that we are not aware of the gap.

A team can be convinced that it values continuous learning while subtly punishing every mistake. It can believe it is self-organizing while waiting for the boss to decide at every important fork in the road. It can claim to deliver value when in reality it delivers features and never checks whether they are useful. Overestimation is usually not lying or cynicism. It is a genuine blindness to the distance between what we profess and what we practice.

Where teams overestimate themselves the most

Although every team is different, there are a few areas where the gap between perceived and actual performance appears most often. They are worth walking through, not as an accusation, but as a mirror.

The first is self-organization. Many teams consider themselves self-organized because they decide small things day to day, but the decisions that really matter are still made outside the team. Autonomy in trivial matters and dependence in essential ones is not self-organization, even if it looks like it from the inside.

The second is value orientation. It is common to believe the team works focused on user value when what is actually being optimized is delivery speed. We measure how much is produced, not how useful what was produced is. The team feels results-oriented while in reality living in an activity-oriented way.

The third is continuous improvement. Almost every team holds retrospectives and therefore believes it improves continuously. But reflecting and changing are two different things. If the same complaints reappear sprint after sprint with different words, the retrospective has become a ritual of venting rather than an engine of improvement, no matter what the team says in its internal survey about improving nonstop.

The fourth is psychological safety. A team can believe it is an open space where anything can be said, while also having topics nobody talks about and voices that systematically stay silent. Amy Edmondson has shown that psychological safety is the foundation of collective learning, and also that it is easy to overestimate, because those with more power tend to perceive more openness than actually exists for everyone else.

Why self-perception fails precisely here

There is a structural reason agility is especially prone to overestimation. Its visible practices are easy to adopt, while its substance is hard to master and hard to measure from the inside.

Starting a Daily Scrum is a matter of a day. Building the trust that makes people tell the truth in that Daily Scrum can take years, and no board shows it. Because the visible is implemented quickly and the essential matures slowly, it is easy to confuse having adopted the forms with having achieved the substance. The team looks at its ceremonies, checks that they are all there, and concludes it is agile, without noticing that the ceremonies are the packaging and not the content.

On top of this, the judge and the party being judged are the same. When a team evaluates its own agility, it does so with the same assumptions and blind spots it wants to assess. It is like trying to read the label from inside the jar. That is why pure self-assessment, without any external reference or objective data, tends to confirm what the team already believes about itself.

How to look at ourselves more honestly

Reducing overestimation does not mean flogging yourself or assuming everything is done wrong. It means introducing mirrors that return a less flattering image than self-perception does.

A first mirror is real behavioral data. Instead of asking whether we are improving, look at whether the same problems keep coming back. Instead of believing we deliver value, check whether anyone uses what we deliver and whether it helps them. Facts are less flattering and more useful than feelings.

A second mirror is an outside perspective. An agile coach, a sister team, or any observer who does not share the group's blind spots can see what is invisible from within. The condition is that this observer feels free to say uncomfortable things and that the team is willing to listen without getting defensive.

A third mirror is diagnostic frameworks that structure reflection with clear references, rather than leaving it to general impression. Along these lines, AgiLevel is Scrum Manager's tool for assessing an organization's agility by looking at both its operational practices and its cultural values, without imposing a specific framework. Its value is not in giving a score, but in making explicit the distance between what the team believes about itself and what its practices reveal, which is exactly the blind spot that overestimation hides.

The AI context: a new layer where we can overestimate ourselves

It is worth adding a recent area where overestimation is already appearing strongly: the adoption of artificial intelligence itself. Many teams have started using AI tools and, as a result, feel they master their integration into their work. Using a tool and mastering it, however, are not the same thing.

A team can believe it uses AI with judgment when in reality it accepts what the machine proposes without reviewing it, attracted by the fluency of a convincing-sounding result. It may think it has gained maturity when it has only gained speed, and sometimes speed for making mistakes faster. The same gap between perception and reality that affects agility is reproduced here, with an added complication: because the terrain is new, there are even fewer references to calibrate against. Asking yourself honestly, "Are we really directing AI with judgment, or are we simply using it?" is one of the most useful conversations a team can have right now.

Overestimation is not a moral flaw in a team but a very understandable human tendency, which is why fighting it is not about blaming anyone, but about building the habit of looking at ourselves in honest mirrors. A mature team is not one that believes it does everything well, but one that knows precisely where it falters and works on it without drama.

Perhaps the most reliable sign of real agility is not confidence, but a certain calibrated humility, the ability to say "we are good at this, and in this other area we deceive ourselves". That sentence, said out loud and without fear, is worth more than any perfect board.

So the question is serious, and it is for you: in which aspect of agility do you think your team overestimates itself the most? And what mirror would help you see it?

References

• Argyris, C. (1990). Overcoming Organizational Defenses: Facilitating Organizational Learning. Allyn & Bacon.
• Denning, S. (2018). The Age of Agile: How Smart Companies Are Transforming the Way Work Gets Done. AMACOM.
• Dunning, D. (2011). The Dunning-Kruger Effect: On Being Ignorant of One's Own Ignorance. Advances in Experimental Social Psychology, 44, 247-296.
• Edmondson, A. C. (2018). The Fearless Organization: Creating Psychological Safety in the Workplace for Learning, Innovation, and Growth. Wiley.
• Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux.
• Kruger, J., & Dunning, D. (1999). Unskilled and Unaware of It: How Difficulties in Recognizing One's Own Incompetence Lead to Inflated Self-Assessments. Journal of Personality and Social Psychology, 77(6), 1121-1134.
• Larsen, D., & Kua, P. (2016). An Agile Adoption and Transformation Survival Guide: Working with Organizational Culture. Leanpub.
• Rigby, D. K., Sutherland, J., & Takeuchi, H. (2016). Embracing Agile. Harvard Business Review, 94(5), 40-50.