3 Comments
User's avatar
Chris Shaffer's avatar

I heard a while back there was some research that was showing the feedback sandwich might not be ideal. https://phys.org/news/2025-08-compliment-sandwich-longer-effective.html is one recent example. You can find a few others with some alternative approaches recommended.

Jeff Bogdan's avatar

I'm not sure "not ideal" is the right summary of that. Thanks for sharing that post. Another one that I have leveraged is https://www.themuse.com/advice/feedback-sandwich. Combining the insights of both of these posts, what I would say for the summary is: (1) if _all_ you do is use the feedback sandwich, the predictability of it will make people cringe at the first piece of positive feedback, knowing it's a setup for the "meat" that will follow, or (2) if you don't take equal care with the bread, then you're not doing the sandwich right. I remember early on with this method I explained the bad way to do a feedback sandwich: "Hey Don, love your hair. Your code is complete trash. Nice shoes, too."

I love the point on "candor" from your link. This relates to the Nietzsche quote that I used ... the long term goal in all of your feedback delivery is in building their threshold for truth tolerance. So the onus is on the deliverer to make it an ultimately positive experience for the recipient. When the receiver comes away from a _constructive_ feedback experience with a _positive_ sentiment, things are going in the right direction.

The feedback sandwich is a tool in my toolbelt. But there are others that I leverage accordingly. Like the much more recent "Yes! That!" which I called out after the sandwich.

Keep the insights coming, Chris. I love your additions!

Chris Shaffer's avatar

Fair. And, I also think I heard certain personality types react better to the sandwich than others, so knowing your audience is also really important for what tools you use.