Leggi questa pagina in italiano
The Feedback Sandwich
What the feedback sandwich is in Giovanni Ceroni's NLP: the three layers for giving effective feedback that opens gates instead of closing them.
Bad feedback shuts down. Good feedback guides. Great feedback lights up what the person can't see for themselves.
What it is
The feedback sandwich is a structured communication strategy for correcting without humiliating, guiding without shutting down, helping someone grow without triggering defensiveness. It's built on one of NLP's fundamental presuppositions: "all results and behaviors are achievements, whether or not they're the desired result for a given task or context" — feedback versus failure. When it's built well, feedback doesn't wound: it builds.
Why it matters
The feedback sandwich matters because it respects the natural sequence through which the brain decides whether to open up or shut down — the Three Gates. If even one of these levels isn't passed, defenses activate and no information can get through, regardless of whether it's true, useful or smart: it simply becomes unusable for the person receiving it.
How it works
Before starting, two preliminary rules apply. The first rule is asking permission: "can I give you some feedback on this?" Unsolicited feedback tends to make the recipient stiffen up, while asking permission lowers resistance and opens up listening. The second rule is congruence: the tone of voice, physiology, timing and emotional state of whoever gives the feedback need to be aligned with the content — even the best content, delivered badly, loses value. Before starting, a pattern interrupt can help, with a generic, light question ("where did you have lunch?"), to avoid starting off in an already charged emotional state.
The structure of the feedback sandwich has three layers:
The first layer — what went well. It's not there to flatter: it's there to open. When the person hears what worked, they feel acknowledged and stop wondering whether they're about to be criticized — this opens the first gate ("friend or foe?"). Being specific matters: "you handled the client well, staying calm and clear" works much better than a generic "good job."
The second layer — what can be improved. It keeps the openness and turns it into usefulness. This is where the distinction between mistake and margin happens: a mistake judges the past, a margin enables the future. When the focus shifts from "what's wrong" to "what can be improved," the mind stops defending and starts evaluating — this opens the second gate ("what's useful for me here?"). It's helpful to say "I noticed that when people interrupt you, you tend to stiffen up and lose your clarity" instead of "you're not capable of handling interruptions": a brain oriented toward improvement responds better than a brain put on trial.
The third layer — how to improve. It gives direction and completes the process: direction doesn't correct, it orients. It shifts attention from what was done to what can be done, opening the third gate ("does this question me, or my beliefs?"). When the focus stays on the behavior and not the identity, the person's worth stays constant, and their beliefs stay flexible — the minimum condition for change. It's helpful to offer concrete, usable guidance: "you can pause, smile and respond calmly, or ask to collect questions at the end of the talk" instead of a generic "work on it."
Every time a person receives information about themselves, the brain runs three quick, automatic checks: intention ("is this against me or for me?"), usefulness ("does this help me or hurt me?"), identity ("does this question me, or just my behavior?"). The feedback sandwich is built precisely to pass, one at a time, these three checks. When things went well and important results were reached, it's also useful to celebrate, with rewards proportionate to the result achieved — and you can even use anchoring if the emotion is particularly strong.
Common mistakes
A common mistake is skipping the request for permission, giving unsolicited feedback that automatically triggers resistance. A second mistake is phrasing the second layer as a judgment of the person ("you're not capable") instead of the specific behavior ("when X happens, Y tends to follow"). A third mistake is closing the feedback without the third layer, leaving the person aware of a margin for improvement but without concrete tools to close the gap.
Practical example
A manager wants to give a team member feedback on a presentation just finished. They ask permission: "would it be okay if I shared a couple of observations about the presentation?" They open with the first layer: "the delivery was clear and the pacing helped follow the thread." They continue with the second: "one thing that could improve is the questions: when you answer right after someone jumps in, there's a slight sense of discomfort from you." They close with the third: "you could let a few seconds of silence pass before answering, to communicate that you're really listening." The team member receives precise, usable information, without feeling judged as a person.
Applications
The feedback sandwich applies to leadership and managing team members, coaching, education and parenting, personal relationships, and any context where an area for improvement needs to be communicated without compromising the relationship or the motivation of the person receiving it.
Frequently asked questions
What is the feedback sandwich?
It's a three-layer communication strategy — what went well, what can be improved, how to improve it — designed to correct a behavior without triggering the defenses of the person receiving the feedback.
Why is it important to ask permission before giving feedback?
Because unsolicited feedback tends to make the recipient stiffen up, triggering resistance. Asking permission, even with a simple question, lowers that resistance and opens up listening.
What's the difference between a mistake and a margin in the second layer of feedback?
A mistake judges the past ("you did it wrong"), while a margin enables the future ("there's room to improve here"). The second kind of language lets the mind evaluate instead of just defending itself.
Why does the third layer of feedback need to be concrete?
Because just pointing out a weak spot, without offering tools or a method, leaves the person aware of the problem but without a real direction for solving it. The third layer needs to orient, not just indicate.
What does the brain automatically check when it receives feedback?
Three things in sequence: intention (against me or for me?), usefulness (does this help or hurt me?) and identity (does this question me, or just my behavior?). The feedback sandwich is built to positively pass each of these checks.
Related concepts
The Three Gates, The Blind Spot, Words and Biochemistry, NLP Presuppositions, Effective Communication.
Go deeper
The feedback sandwich, with complete applicable examples, is presented in the chapter of the same name in Volume I of "The Invisible Blade".
Go deeper in the books
If this topic is useful to you, you can explore it further in the "The Invisible Blade" series, where concepts are connected to examples, models and practical applications.

