FORWARDED THIS? Subscribe here to get the next one.

THE PERMISSION GAP

On Thursday I sat on a panel talking about AI and compensation and the question of what you’re allowed to do vs what you want to do came up.

Coincidentally, a month ago I launched a survey called the “permission gap” precisely to shed some light on something I’ve been hearing a lot inside Range. Something along the lines of “at my company we’re only allowed to do X, but I want to get the team to do Y through Z”.

I want to share two findings from the permission gap report which may resonate with you.

More than 1 in 2 respondents so far are being pushed to adopt AI and denied the tools to do it at the same time.

(Are you serious?)

Yes, and almost everyone I speak to about it assumes they're the only one.

A licence doesn't fix it.
Nearly 3 in 4 of the people on a full enterprise licence still can't do AI work on real compensation data because something else is in the way. Usually they can't install anything, or their approved tool is chat only, or comp has been carved out of a policy that allows AI everywhere else.

What I witnessed by speaking to reward leaders inside and outside of the community is that when people can't experiment at work they experiment at home instead, which moves the risk somewhere the employer can't see it.

The fix isn't that complicated. Give people the tools, a safe environment where to experiment, realistic data that isn't real, and clear guidelines about what they can do with it.

The driver behind all of this is that your IT is trying to catch up as much as you are, and they feel behind the AI curve as much as you do. So today I decided the show must go on.

I’ve reopened the benchmark again so you can continue monitoring how stuck or free you are compared to other comp and total rewards teams without imposing a deadline.

I’m going a step further. I want to help you get unstuck so I’m also sharing with you the dataset I use to experiment with AI in Comp safely.

What would your employer have to change for you to do this work properly? Reply and tell me. It decides what the next benchmark asks.

Last Sunday: Issue 13. The comp build that starts by throwing half the requirements away

Most of us are the only comp person in the building working this out alone. If you know another one, forward this issue and point them at your link below. Once three of them subscribe, the next Comp AI guide reaches you a week early.