I was sitting in church this weekend when my pastor asked a question that stopped me mid-note-taking.
“When you’re about to get good news and bad news,” he said, “which do you want first?”
He asked the congregation to raise their hands. Overwhelmingly, people wanted the bad news first. Then he made a claim that I haven’t stopped thinking about since: bad news gets our attention faster, and it moves us to decisions faster.
The first half of that, it turns out, is backed by solid research. The second half, the part about decisions, is more complicated than it sounds, and the complication is exactly what I want to talk about, because it’s the same complication I’ve been coaching my improv trainers and show directors through for years.
Why Negative Feedback Gets Our Attention Faster
There’s real science behind why bad news captures us. Researchers call it negativity bias, and it’s not a personality quirk; it’s wiring. Your amygdala reacts to threatening information in roughly 100-120 milliseconds, before your visual cortex has even finished processing what you’re looking at. Your brain is built to notice the bad thing first. Negative information consistently gets weighted two to five times heavier than positive information across attention, memory, and decision-making. “Bad is stronger than good” is a documented pattern.
So when the room said “give me the bad news first,” they were behaving exactly the way brains are built to behave.
Why Positive Endings Can Reduce Accountability
A well-known study by psychologists Angela Legg and Kate Sweeny found that people do prefer bad news first; somewhere between 78 and 88% of us depend on it. But when they tracked what actually happened after people received the news, they found something that cuts against how most of us lead.
People who heard the bad news first ended up in a better mood. The good news that followed took the edge off, and they left the conversation feeling relieved. But that relief came at a cost: they were less motivated to actually change anything.
People who heard the bad news last stayed unsettled. That discomfort didn’t go away just because the conversation ended. And that lingering unease was exactly what pushed them to act … to watch the training video, to make the change, to follow through.
In other words: leading with bad news makes people feel better. Ending with it is what makes people do something.
That’s the opposite of how most of us were taught to deliver feedback. We were told to “sandwich” the hard stuff … soften the landing, end on encouragement, leave them smiling. It turns out that instinct, the one that feels the most compassionate in the moment, may be the exact thing that lets people off the hook.
The Feedback Sandwich
If you’ve managed people for any length of time, you’ve probably used this method even if you never knew it had a name. The feedback sandwich is exactly what it sounds like: you open with something positive, slide the actual criticism into the middle, and close with another positive to send the person off feeling good.
It became popular for good reason. It’s kind. It softens a moment that can otherwise feel confrontational, and it protects the relationship between the person giving the note and the person receiving it. For decades, it’s been the default advice in management training, coaching certifications, and performance review templates everywhere.
The problem is that the same research showing bad news captures our attention faster is also starting to show that the sandwich may work against the very outcome we’re using it for. When the hard truth gets buried in the middle, and the conversation ends on praise, the recipient often leaves comforted, but comforted isn’t the same as motivated. Newer research suggests the sandwich may be optimized for how the moment feels, not for what happens after the moment ends.
Why Most Leaders Avoid Honest Feedback
I’ve spent years training improv performers, and even longer training the trainers and directors who coach them. This is the single hardest skill I have ever tried to teach: how to give feedback that actually changes behavior.
Here’s what I continuously see: A new director sits down after a rehearsal or a show, and they know a scene didn’t work; the performer missed an obvious “yes-and,” they blocked their partner’s offer, they played it safe instead of making a bold choice. Instead of naming it clearly, the director buries it. They open with three compliments, slip in one soft, hedged piece of critique, and close with “but overall, great job tonight!”
The performer walks away feeling good. And nothing changes.
I understand exactly why leaders do this, because I’ve done it myself. Giving direct feedback feels risky. You worry about deflating someone, damaging the relationship, being “that” leader. Instead, you protect their feelings and your own comfort by leading with positive, burying the real note, and closing warm.
What the research names so clearly is something I’ve always sensed but never had the language for: that instinct to end on a high note is often more about the giver’s discomfort than the receiver’s growth. Legg and Sweeny call this an egocentric bias: news-givers tend to choose the order that feels better to deliver, not the order that actually serves the person hearing it.
That’s exactly what I witness with my improv comedy show directors. After the shows, during the debrief sessions, directors avoid feedback, or dilute it into something so soft it doesn’t land, because the discomfort belongs to them in that moment, not to the performer who needed the note.
How to Give Feedback That Leads to Real Behavior Change
This reality has reshaped how I train the trainers at Carolina Improv Company. I’m not telling anyone to abandon warmth or start every note with a gut punch. Instead, I am telling them the sandwich has a flaw: if you end on the positive, you may be giving your performer permission to feel good and move on, instead of giving them a reason to actually work on the thing that needs work.
I built a framework called Improvised Intelligence™, which applies the principles of improv to leadership and organizational development. One of its core ideas is what I call the New Choice Effect: growth requires a moment of real discomfort before a shift can happen. You can’t make a new choice, a better one, if nothing ever pointed you toward needing one. A feedback sandwich that ends warm can accidentally erase that moment. The discomfort gets absorbed before it has a chance to do its job.
Here’s what I’m asking my leaders and trainers to try: name the real note clearly and let it be the last thing the performer carries out of the room. Not cruelly. Not without context or care. But without smoothing it over with a closing compliment that lets everyone off easy.
It’s uncomfortable. It should be. That discomfort is doing something the polite version never could.
Key Takeaways for Managers and Leaders
- Bad news naturally captures attention; your brain is wired to notice it first.
- Ending a conversation with praise can feel kind, but it can also quietly reduce someone’s motivation to change.
- Honest feedback builds trust when it’s delivered with care, not when it’s diluted to avoid discomfort.
- Discomfort isn’t failure. It’s often the beginning of growth.
- The goal of feedback isn’t to make someone feel better. It’s to help them get better.
Should You Ever End Feedback on a Positive Note?
My pastor’s larger point still stands, even with the nuance the research adds: bad news matters, and how we handle it shapes what happens next. The order isn’t just about softening the blow for the person receiving it; it’s about whether we’re willing to let the discomfort do its work, instead of rushing to relieve it because it’s uncomfortable for us to hold.
Next time you’re about to give someone feedback, ask yourself the honest question: are you ending on a high note for their benefit, or for yours?
Sources: Legg, A. M., & Sweeny, K. (2014). Do you want the good news or the bad news first? The nature and consequences of news order preferences. Personality and Social Psychology Bulletin, 40(3), 279-288.
