Why fair, specific, timely feedback is one of the highest-leverage practices in a technology organization.
In engineering organizations, feedback is often treated as a soft skill. That is a mistake.
Feedback is an operating mechanism. It helps people understand the gap between intent and impact, gives teams a chance to correct course, and makes career growth less dependent on guesswork. In the TalkingTech Leadership OS, this belongs to People & Culture because healthy organizations do not rely on talent alone. They create the conditions in which talent can see reality, learn quickly, and take on more responsibility.
The standard is simple: feedback should be fair, assertive, specific, and connected to improvement. It should describe the situation and the observable facts, not make a judgment about the person. It should happen soon enough for the context to be clear and the correction to still matter.
Feedback is a control loop, not a verdict
A performance conversation is not a declaration of what someone is. It is information about what happened, what effect it had, and what should happen next.
That distinction changes the language. “You are careless” is a judgment about identity. “In yesterday’s deployment review, the rollback plan was not documented, so the on-call engineer had no tested recovery path” is a description of a situation, a behavior, and an impact.
The second statement is more useful because it can be discussed and acted on. The person can explain the context. The manager can clarify the expectation. The team can agree on a concrete change, such as adding rollback readiness to the release checklist or reviewing it during the change approval process.
This is the logic behind the Situation–Behavior–Impact model recommended by the Center for Creative Leadership: identify the specific situation, describe the observable behavior, and explain the impact without inserting opinions or labels. The model can be used for corrective feedback and recognition alike. 2
The goal is not to make feedback cold. The goal is to make it accurate enough to be fair.
Why speed matters
Feedback has an expiration date.
When a manager waits several weeks to mention a problem, three costs accumulate. The person has more time to repeat the behavior. The team has more time to adapt around it or absorb its consequences. The original context also becomes less precise for everyone involved.
That delay creates a second problem: feedback becomes heavier. A small course correction that could have taken five minutes becomes a formal conversation filled with accumulated examples, frustration, and surprise.
Fast feedback does not mean impulsive feedback. It means close enough to the event that the facts are still available, delivered with enough thought to be useful, and separated from the emotional heat of the moment. For a routine issue, “later today” is often better than “at the next performance review.” For a serious incident, the leader may first stabilize the situation and then return to the behavior when both sides can think clearly.
Gallup describes meaningful feedback as timely, relevant, focused, actionable, and future-oriented. Its central point is practical: employees should leave the conversation knowing what to keep doing, what to change, and how the change will improve the work. 3
The damage caused by silence
Not giving feedback is not neutral. It is a management decision with consequences.
When a manager sees a recurring issue and says nothing, the employee may reasonably conclude that the behavior is acceptable. The organization then pays for the gap through missed outcomes, rework, customer friction, operational risk, or increased load on stronger teammates.
The employee pays as well. Without clear information, they lose the chance to practice a better technical approach, improve communication, repair trust, or prepare for a larger role. They may discover the problem later through a denied promotion, a difficult performance review, or an unexpected exit conversation. At that point, the feedback feels like a hidden rule that was enforced retroactively.
Silence also damages fairness. The people who receive direct coaching get a chance to improve. The people who receive vague reassurance do not. If leaders protect someone from a difficult conversation, they may be protecting the short-term comfort of the manager while reducing the employee’s long-term options.
False feedback is more harmful than uncomfortable feedback
Feedback that is not true is not kind.
There are two common forms. The first is empty praise: telling someone they are doing well when the evidence shows a meaningful gap. The second is softened criticism: using vague language such as “be more strategic” or “improve your communication” without naming the observable behavior that needs to change.
Both forms create uncertainty. The employee cannot tell which strengths are real, which behaviors are limiting their impact, or what evidence will show progress. Managers then become frustrated because the person has not improved, even though the organization never gave them a usable description of the problem.
Truthful feedback does not require harshness. It requires a clear connection between evidence, expectation, impact, and support. If the concern is technical, name the technical gap and the expected standard. If the concern is behavioral, describe what was said or done and how it affected the team. If the issue involves both, separate the two so that the employee knows which kind of practice is needed.
A useful test is this: could the employee turn the feedback into a specific action this week? If not, the message is probably too vague.
A practical structure for fair feedback
A concise feedback conversation can follow six steps.
1. State the purpose
Open with the reason for the conversation. “I want to review what happened in the incident handoff and agree on how we can make the next one safer” is clearer than “We need to talk about your performance.”
2. Describe the situation
Name when and where the relevant event occurred. Avoid assembling a general case from every frustration of the past year. If there are multiple issues, handle them as separate conversations or explain the pattern with evidence.
3. Describe the behavior
Use observable facts. Avoid adjectives such as careless, difficult, weak, uncommitted, or not strategic. These labels invite a debate about personality instead of a discussion about choices.
Instead of saying, “You were dismissive in the design review,” say, “When Ana raised the data-retention risk, you interrupted twice and moved the discussion to implementation before the concern was answered.”
4. Explain the impact
Connect the behavior to a result. The impact may affect delivery, reliability, customer trust, decision quality, team participation, or the person’s own credibility.
“Because the concern was not resolved, the team approved an approach that required rework two days later” gives the employee something concrete to understand.
5. Ask for context and intent
Feedback should be a conversation, not a prosecution. Ask, “How did you see the situation?” or “What was happening from your perspective?” This does not erase the impact. It helps distinguish intent from effect and may reveal a system problem, missing context, or a constraint the manager did not see.
Psychological safety is not the absence of accountability. It is the condition that allows people to report mistakes, ask for help, and disagree without fear of retaliation. Leaders build it by inviting input, listening without punishing bad news, and making learning from errors a normal part of work. 4
6. Agree on the next behavior and follow-up
End with a specific change. Define what good looks like, what support is available, and when you will revisit the topic. A conversation without a next step is only an observation.
For example: “In the next two design reviews, pause after each risk is raised and summarize the concern before proposing a solution. I will join the first one, and we will review what changed on Friday.”
Feedback should recognize leverage, not only correct problems
A feedback culture that only appears when something goes wrong will feel punitive. Recognition matters, but it should be as specific as corrective feedback.
“Great job” is pleasant but weak. “In the incident review, you separated the facts from assumptions, made the customer impact visible, and helped the team agree on two prevention actions” tells the person which behaviors created value. That information can be repeated in a harder situation and can become evidence of readiness for broader responsibility.
The same principle applies to senior people. Leaders should receive feedback about the conditions they create: whether decisions are clear, whether dissent is welcomed, whether priorities are stable, and whether teams are being asked to absorb avoidable complexity. Feedback is not only for fixing individual contributors. It is one of the ways an organization improves its operating model.
Feedback is the engine of career growth
Career growth becomes clearer when feedback is connected to capability rather than personality.
A career conversation should answer four questions:
- What outcomes is the person already producing consistently?
- Which behaviors or skills limit the scale of their impact?
- What evidence would show that they are ready for the next level?
- What opportunity will allow them to practice that capability?
For a senior engineer, the opportunity may be leading an architecture decision across teams. For an engineering manager, it may be improving planning quality, developing another manager, or making trade-offs visible to executives. For a tech lead, it may be moving from solving every technical problem personally to creating an environment in which the squad can solve problems safely without them.
This is why feedback should not be reduced to a rating. A rating summarizes a moment. Development requires a loop: expectation, observation, feedback, practice, and follow-up. The loop becomes stronger when the employee helps define the goal and chooses part of the learning plan.
The manager’s role is to make the path visible. The employee’s role is to engage with the evidence and practice the change. The organization’s role is to provide real opportunities, not just advice.
Build a feedback operating system
A healthy feedback culture is designed through repeated habits.
Start by defining a simple feedback contract for the team: discuss behavior, not identity; raise material issues early; be specific; ask for context; protect confidentiality when correction is sensitive; and close the loop. Leaders should model the contract by asking for feedback on their own decisions and by responding without retaliation.
Use different settings for different purposes. Give recognition publicly when it helps the team learn what good looks like. Give corrective feedback privately unless the issue is a shared process failure that the group needs to understand. Do not use a public meeting to make an individual example of someone.
Create regular one-on-ones where feedback is expected, not exceptional. Use performance reviews to synthesize patterns and career direction, not to reveal months of unspoken concerns. Document agreed actions briefly so that memory does not become the only operating system.
Finally, calibrate for fairness. Ask whether the same behavior would have been interpreted differently for another person, whether expectations were explicit, and whether the employee had the authority, information, and support required to meet them. Feedback cannot correct a system that is itself unclear.
The standard
The standard for feedback is not “nice” or “tough.” It is true, useful, timely, and fair.
True feedback respects the person enough to describe reality. Useful feedback explains what to do next. Timely feedback reduces the cost of correction. Fair feedback focuses on evidence and expectations rather than labels.
When leaders work this way, feedback stops being an annual event or a manager’s personal style. It becomes part of how the organization learns, how teams protect results, and how people build the capabilities that move their careers forward.
That is the People & Culture pillar in practice: high-trust teams with clear standards, direct conversations, and enough psychological safety to improve before small problems become organizational ones.
Top comments (0)