Compress your feedback like a data file

How to help people hear what you’re trying to say

I was giving a founder some feedback recently and realised they’d gone into fight or flight mode. What they were hearing and what I was saying had started to diverge wildly. I empathised because I’ve been there — it can be really tough to hear tough feedback. So I put some thought into how we can “compress” feedback and reduce the amount of emotional bandwidth required to receive it.

What is emotional bandwidth? It’s the energy required to manage your emotional state and stay present, open and able to listen. Some conversations don’t require much of this bandwidth — feedback conversations generally do.

So how do you give tough or important feedback while minimizing the emotional bandwidth required by the other person? You compress it. Our favorite internet content comes to us compressed. The names of the compression algorithms have made their way into how we talk: gifs, jpgs, mp3s.

Compression involves using (1) processing power upfront and (2) a clever algorithm to squash content to a size that can fit through the narrow pipe of bandwidth going to our phones or homes.

When content exceeds the available bandwidth your message is not received. E.g. you get a frozen video stream. Exceeding someone’s emotional bandwidth has basically the same effect: they end up in fight or flight and unable to internalize what you’re saying.

Even if you’re not exceeding available bandwidth, not compressing things means there’s less room overall. Less simultaneous Netflix and Spotify streams. Less ability to be present with family, friends and colleagues. Less ability to capture information, process and reflect.

So how do you compress feedback? Remember compression requires (1) applying processing power upfront and (2) a clever algorithm. Applying processing power upfront to your feedback means you need to actually write it out. This means you can evaluate its uncompressed “file size”.

Here’s how you assess the file size of your feedback. First, look at how specific or general your observations of the other person’s behavior are. Are you making general statements about their behaviors, attitude or beliefs or talking about a specific action you’ve observed?

Second, look at the level of “ownership” you’re taking over your interpretations of this behavior and the thoughts and feelings it has triggered in you — i.e. are you presenting them as your point of view or as universal truths?

The more general your feedback and the less you “own” it the bigger the file size. As the file size grows, more emotional bandwidth is needed from your recipient. See this highly technical diagram:

Why is this the case? Well let’s think about most feedback. We tend to give general observations (“you don’t seem to be putting enough effort in”). We elevate our own perspective to the status of the truth (“you don’t care about this project”). We also avoid being vulnerable by sharing real emotions (sadness, joy, anger etc) and so substitute judgments for feelings “I feel like you need to make a change”.

Notice how this feedback feels like a universal assessment? Being universal means the recipient needs to either fully accept your take on the world or argue to protect their own perspective. Either way they’re burning up emotional bandwidth.

Compressing feedback means being less universal by moving along two axes: specificity and ownership. Talking about specific actions at specific times allows for times when the person might have acted differently. Talking from just your own perspective gives room for the recipient (and perhaps others) to see things differently.

The result is feedback that doesn’t require the recipient to fully take or reject your perspective. It can stand apart from them. It’s useful data not a universal judgment. So they don’t need as much emotional bandwidth to hear it. Extra points if you’ve been vulnerable by stating (and owning) your actual emotions (sadness, joy, anger etc) — that seems to temporarily increase emotional bandwidth.

Here are some examples:

So what’s the compression algorithm for tough feedback? Here’s my 1.0 version — interested to hear your thoughts:

If you’ve read this far — let me know what you think! What was the best-delivered feedback you’ve ever been given? What was the worst?