Scripts and frameworks for giving hard feedback as a leader

Executive overview

Most leaders avoid difficult conversations because they fear upsetting people — but withholding feedback actively harms the person and the company. The job of a leader is not to make employees happy; it is to drive results and help people grow.

Alisa Cohn provides specific scripts for performance feedback, defensiveness, promotion denials, and firing — plus a founder co-founder alignment framework and a three-question meeting close.

Giving hard feedback without a script is the real risk — preparation makes difficult conversations not just easier, but better for everyone.

Performance feedback scripts

• Open with observable facts, not judgments: "What I've observed is..." or "What I'm hearing from peers is..."
• Establish shared stakes: "We both know the most important way you can succeed here is..."
• End with a clear outcome: state what the conversation needs to resolve before you leave the room
• Tone should be even and matter-of-fact — this is Tuesday, not a crisis
• Build a feedback relationship first: give specific positive feedback regularly so people receive hard feedback through the lens of "she wants to help me"
• Specific positive feedback has the same standard as negative — name exactly what worked and why

Handling defensiveness

• Pause before reacting: "Let's just pause for a second — I can feel the energy has changed"
• Restate intent: "I have no intention of upsetting you; I want to talk about things that will help your career"
• Offer to reschedule if emotion is too high, but make clear the conversation must happen
• Stay grounded in your evidence — you prepared, you're unlikely to be simply wrong
• Frame it as a conversation: "If you have a different point of view, I want to hear it — but we can't keep going without resolving this"

Delivering a promotion denial

• Don't bury the lead — state the decision early, then explain the reasoning
• Give specific reasons tied to what the role requires, not a verdict on the person
• Paint hope for the future: commit to helping them build skills and advance
• Acknowledge their disappointment directly before moving to rationale
• Offer a follow-up conversation once they've had time to digest the news
• If they push back with irrelevant points (tenure, internal candidacy), mirror their concern back and explain why those factors weren't part of the decision

The pre-firing conversation

• Before firing, have the explicit warning conversation first — this is essential, not optional
• Script: "I've talked to you multiple times about X. After six months, the problem persists. I need you to fix this within 30 days, or we will part ways."
• Name it a difficult conversation at the start so they understand the stakes immediately
• Acknowledge their strengths while being clear these specific behaviours are deal-breakers
• When the firing does happen, it should not be a surprise — the script is short: "We made the decision. We've discussed this multiple times. Here's HR."

Your job as a leader is not to make people happy

• Leaders who optimise for happiness avoid hard conversations, allow underperformance, and drift toward cultural decay
• Avoidance leads to a demoralising environment for high performers who see the problem going unaddressed
• High engagement comes from a winning culture: clear roles, known impact, celebrated milestones
• The right people — those who like to win — will join and stay in a results-driven culture
• Founders especially conflate "people like working here" with "we are set up to succeed" — these are different conditions
• Culture is not perks and socials; it is "how we get work done around here"

Three questions to end every meeting

• "What did we decide here?" — ask everyone; six people in the same meeting often give six different answers
• "Who needs to do what by when?" — explicit owners and deadlines, not implied action items
• "Who else needs to know?" — decisions often never cascade; cascading is part of the decision
• Anyone in the meeting can introduce this ritual — you do not need to be the most senior person in the room
• Leave five to ten minutes at the end of every meeting specifically for these three questions
• Designate a person who enjoys follow-up to capture and distribute answers

The founder prenup

• 65% of startups fail due to founder conflict — most of it was preventable
• Values alignment: each founder picks their three to five core values independently, then compares; adjacent values are fine, conflicting ones need explicit discussion
• Vision of success: "What does this company look like when it reaches its full potential?" — venture scale vs. lifestyle business diverges quietly until it's too late to resolve
• Conflict handling style: how you think you handle conflict vs. how people close to you say you handle it are often different; know the gap before you co-found
• Decision process when you disagree: agree upfront on a method — most expertise wins, most passion wins, or alternating — so disagreement has a resolution path
• Company culture: what kind of environment do you want to build? Results-focused and family-feeling can coexist, but co-founders pushing in opposite directions create two companies inside one

Personal operating manual

• A lighter version of the founder prenup for any team relationship
• Covers: preferred communication style, ideal working blocks, best way to reach you for urgent items, pet peeves, delegation preference (check in regularly vs. report when done), and how to earn a gold star
• Run as a team activity — nobody knows your operating style until you tell them
• Reduces low-stakes conflict so energy is saved for disagreements that actually matter

Patience vs. process

• When a project stalls, leaders default to "have patience" — often the real issue is a broken process
• Signal it is a process problem: you cannot picture how this comes together, you have not talked to the people involved, there is uncomfortable silence around the topic
• Fix: get closer to the work, talk to the people, look at the data — the blocker is often one small thing that needs to be unstuck
• Hope is not a plan; a visible path forward is