What if your best mentor only had to explain things once?
Ask a new hire where they go when they’re stuck, and you’ll almost always hear a coworker’s name before you hear about a “how to” deck or a shared folder. Usually, it’s the person who knows why the partner deck is built the way it is, which co-marketing requests always come back with the same edits, and what the client actually meant when they said, “make it pop.”
Nobody hired that person to be the team’s memory. It just happened, one “quick question” at a time. Meanwhile, the newer folks asking those questions are sharp and eager, and they’re spending a surprising part of their week chasing things that live in someone else’s head. Neither side is doing anything wrong. It’s a system that depends on sitting near the right people, and it gets harder to maintain as a team grows.
How Most of Us Learned the Job
Think about how you learned your job. Probably not from the onboarding deck. More likely, you learned by sitting in on a call you weren’t really needed for, turning in a draft that came back covered in edits, and working up the nerve to ask why. The grunt work wasn’t glamorous, but it kept junior people close enough to senior people their judgment was passed down along the way.
That kind of learning is getting harder to come by. AI is very good at the tasks we used to hand off to new hires: first drafts, research summaries, formatting, and the tenth version of the same email. Some companies are already hiring fewer entry-level people because of it.1 Even where those roles stay, a lot of the work that taught people the business is now being handed to AI.
Add remote and hybrid work — which we’re big fans of — and there’s no hallway left to overhear things in. Most of what people used to pick up by accident now has to be planned.
What Walks Out the Door
At the same time, the people who carry that knowledge won’t be around forever. They get promoted, switch teams, move on, or retire, and most of what makes them good at their jobs was never written down.
You won’t find it in a shared folder. It’s knowing which partner needs a phone call instead of an email or why the messaging framework has that one line nobody’s allowed to cut. When that person leaves, the rest of the team has to relearn it all, usually the hard way.
Where AI Helps
The same technology that’s shrinking the entry-level workload is also pretty good at passing expertise along. When Stanford and MIT researchers gave customer support agents an AI assistant that learned from how their best colleagues handled problems, the newest agents resolved 34% more issues per hour, while the veterans saw little change.2 The assistant didn’t know anything the top performers didn’t. It made what they knew available to everyone else, right when they needed it.
That’s where we see the real opportunity for most teams: using AI to make mentorship go further. A senior person explains their thinking once, and it’s there for the next new hire and the one after that. The new hire gets an answer at 4:45 on a Friday without feeling like they’re bothering anyone, and the mentor gets time back for the conversations that need a human.
Making It Work in Practice
Capture the reasoning behind the work. Most teams already have process docs. What they’re missing is the why: why a campaign was structured a certain way, why one partner got a different approach, why a draft was sent back. Record a quick walkthrough after a big project. Keep a running decision log. Ask your senior people to talk through their edits instead of just making them. That’s what an AI tool needs to learn from, and plenty of it already exists in Teams chats and meeting recordings. It just isn’t organized in a way anyone can use yet.
Let your experts be curators instead of help desks. Nobody enjoys answering the same question for the fifth time. When AI handles the repeats, senior people can shift to reviewing what the system says, correcting it, and filling in what’s missing.
Rethink entry-level work instead of cutting it. The new version might be less “write the first draft” and more “take the AI draft and make it better.” That last part is where the learning happens, and it keeps a real person in the loop.
Finally, leave the personal side of mentoring to people. A knowledge base can tell a new hire what the team knows. It can’t tell them they did a great job, pull them into the room for a big client moment, or notice when they’re struggling. Those still take a mentor who’s paying attention.
Capture It Once
We think about this a lot, because it’s the same idea behind how we work with clients. When we build a messaging framework in ContentGen®, our strategists do that thinking once, and every banner, one-pager, and email in the campaign draws from it. A team’s expertise can work the same way. Once it’s captured well, it doesn’t depend on one person being available.
Your senior people will still be the ones new hires look up to. With a little planning, they won’t have to be the only place the answers live.
Sources
1AI is wiping out entry-level jobs. Here’s how to surf the wave and not get crushed by it | Fast Company
2Generative AI at Work | National Bureau of Economic Research
