FORWARDED THIS? Subscribe here to get the next one.
Run your next comp proposal past an AI committee first
Yvonne Prang led global rewards at McDonald's for over 90,000 employees across 16 countries, and is now VP, Group Rewards & Mobility at Swissport. She also keeps a digital board, and she walked me through it on the Range Podcast. It's a team of AI agents, each built with particular skills and a knowledge base, and she gives them names because it makes a proper conversation easier.
Meet the board
Selena is the strategist. Max is the M&A specialist, and flags things like an EBITDA multiple outside the usual range. Frank Finance is her CFO, the one she asks whether free cash flow is enough. Carmen is her Chief People Officer, who asks her questions back. Between them the agents tell her which question she hasn't answered and which angle she hasn't looked at, and she says the work gets better with each loop.
She pushes back on them. She tells an agent she doesn't believe something, asks where it got that and what the reasoning is, and describes the board as more a challenger team than one that does what she tells it. Something that used to take her days now takes hours, or even still a day, and she says the quality is much better. She isn't sure how much time that really saves, because she puts it back into cross-checking and keeping her AI literacy up.
Why she trusts it
I asked her what makes a board like that reliable enough to take to a real board. She tells each agent what good looks like and what not to do, which she thinks is the more important part. She asks for a confidence level every time, and for a list of sources that she checks herself. She also cross-challenges, pulling in a second agent to verify what the first one did.
Each agent has knowledge documents she loaded, and she refines them often. When an answer isn't what she wanted, she gives one of her agents the current document or prompt along with the problem and asks how to improve it. She also keeps her own judgement in the loop. She says she knows what good looks like, and when something doesn't feel right she may go and research it herself in Perplexity before challenging the agent.
Yvonne points her board at strategy, deals and finance questions. The decision on your desk is a comp proposal heading for committee, where the CFO, counsel and the chair are all looking for the hole.
The Comp Committee Challenge Board gives you five of those seats to run it past first. They're the CFO, the CHRO, employment counsel, the employee voice and the RemCo chair. Each attacks your proposal from its own seat and comes back with its top three objections, the one question it can't answer, and the evidence that would change its mind. Paste a short summary of the proposal and one seat's prompt into whatever AI you already have open, or paste all five and ask it to flag where the seats disagree.
ON THE RADAR
The pay premium for AI-exposed work has faded to roughly zero. Published 1 October. A US workforce data tracker, updated monthly from job postings and employee profiles, puts the salary premium for AI-exposed work at roughly zero, down from around 2% before ChatGPT. At firms adopting AI, senior headcount shows a 32% relative increase against 6% for junior roles. Issue 15 had Indeed showing AI-exposed pay climbing, mostly at the top. Put the two together and seniority looks like it's doing most of the work. Read → the September tracker
Economists expect AI to press harder on college graduates' pay. Published 24 September. Indeed Hiring Lab's Q3 survey of 123 US academics and economists has its index of AI's wage impact on college-educated workers at 42.4, down from 47.3 last quarter, where anything below 50 leans negative. For non-college workers it's unchanged at 50.9. That's opinion with no pay data behind it, and it points the same way as the Census graduate pay paper in Issue 16. Read → the Indeed survey
One in seven US job seekers would take lower pay for a job that comes with AI training. Published 10 September. A survey of 1,000 US job seekers found 14% would accept less pay in exchange for AI training, and 42% said an employer offering it would be more attractive than a similar one that didn't. That's a benefit with a price on it. Has anyone put it in an offer yet? Read → the survey
RECENT EDITIONS
Grab the free resources
Everything free we've made now lives in one place, the Comp AI Library. The guides, the prompts and the frameworks, all in the open.
Browse the Comp AI Library →
How blocked are you from using AI at work?
Find out in 3 minutes. The permission gap is an ongoing survey that not only gives you a score. It tells you where you sit against other comp and benefits teams that have already participated to the study.
How do I compare against other total rewards teams? →
Join other 2,000+ compensation and total rewards professionals. Get Range Weekly in your own inbox every week. Subscribe free |

