😶‍🌫️ HR built performance management. Managers used it to hide

TL;DR

😶‍🌫️ HR built performance management. Managers used it to hide

👂 How exceptional managers use AI differently to manage performance

🎧 Listen to the newsletter here

This newsletter edition is brought to you by Rippling 💛

😶‍🌫️ HR built performance management. Managers used it to hide

Good question. Wrong problem😅. Let’s take a step back.

Being a great manager is a tremendous amount of work.

Prepare for the 1:1. Listen intently while simultaneously thinking about the right coaching question. Take notes. Summarise afterwards. Set individual goals. And if the manager is genuinely strong, those goals come from observations built across months, not a single conversation. Capture feedback from multiple stakeholders. Write it up carefully. Think about how to deliver it verbally. Then, when review season arrives: revisit every 1:1, analyse the performance trend, write the formal review, and plan how the conversation needs to land.

Are you tired just reading that? 😅 (And we know that's not even half of it!)

Every single action exists in service of 3 things: 

Performance management was built to solve a 2-layer problem and only fixed 1 layer

Managing performance well requires 2 fundamentally different things. I call this The 2-Layer Problem.

For decades, HR built systems to solve Layer 1. We called it performance management. What we might not have admitted is that we hoped Layer 1 would somehow carry Layer 2 as well…it didn't.

I've worked with exceptional managers. The great ones put in e.v.e.r.y. bit of that effort, and it costs them significant time. The ones with good intentions but not enough bandwidth? The system becomes governance: “HR told us to do it, so we did.” The ones who fundamentally don't care? They ignore it anyway. 🤷🏻‍♀️

So in HR, we built structures to compensate. 1:1 templates. Feedback tools. Goal-tracking frameworks. Mid-year check-ins. The theory was sound: build the right system, documentation follows, better management follows.

But what we actually built was a paper trail that let managers say: "We completed the process." Not: "We were great managers."

And that's why we've been "solving" performance management for decades…and it still feels unsolved. 

AI is solving Layer 1 and exposing Layer 2

The managers I know using AI are using it to solve the cognitive constraint, such as tracking trends, surfacing patterns, and capturing what our brain naturally drops in favour of the most recent or most “memorable” moment.

With AI handling Layer 1, a manager can walk into a 1:1 and actually think about the person in front of them. Notice a growth trend across 8 weeks, rather than reacting to whatever happened most recently. Have the conversation about the why behind a pattern. Give recognition that feels specific and earned. Give feedback grounded in actual evidence.

And here's where it gets uncomfortable for managers – and for us. 

When AI removes the cognitive constraint, and the manager still isn't having good conversations, there's nowhere left to hide. Not behind the template. Not behind "we followed the system." For the first time, we'll see clearly whether Layer 1 was ever the real problem. Or whether it was always Layer 2 all along: the capability gap we never truly developed in managers.

The performance management design challenge facing HR

AI is now solving the cognitive constraint, the Layer 1 problem that's been consuming HR's attention for decades. That clears the path to Layer 2: the question we actually wanted to answer all along: do our managers know how to understand, develop, and get the best out of their people?

That's always been the real problem. We just never had the chance to solve it properly.

How exceptional managers use AI differently to manage performance

Layer 1️⃣ in action: How Lukasz solved the cognitive constraint

Lukasz has built a system. Not a complicated one, but a deliberate one.

Every 1:1, he uses AI to transcribe the conversation. But he doesn't use the transcription as the note. He takes his own notes alongside it, such as keywords, phrases, moments that stood out and feeds those highlights back to the AI as a lens.

"If you take that list and give it to AI with the full transcription, it will summarise in a much more meaningful way because it knows what to look for."

Before every 1:1, his system pulls the last session's summary, flags any agreed actions, and surfaces the trend across 8 weeks, sometimes longer.

He also tracks something he calls a Tem’s Sustainability Score, a simple 1 to 4 scale he asks every team member at the start of each session. 

"4 means I could do this indefinitely. 1 means I'm pretty much burned out. The score itself, I don't care about that much. What I care about is when I ask them to explain why."

That question ‘why’ is Layer 2. The conversation that moves someone forward. But it only happens consistently because Layer 1 is handled. The pattern is surfaced and the context is ready. The manager arrives with evidence, not recollection from memory.

Layer 2️⃣ in action: The managerial capability that was always the point

For feedback, Lukasz goes further. He works with each team member to design their own feedback questions, tailored to their development plan, their strengths, and the specific areas they want to grow. Not a standard form and definitely not the same 5 questions for everyone.

For one senior developer, Szymon, he sends the feedback request on Szymon's behalf. Because Szymon had noticed something: when he asked for feedback himself, people were warm, encouraging, but not candid. The authority bias was too strong.

When Lukasz sends it (anonymised, framed as developmental), the responses change entirely.

"I can already see it's getting much more candid. Most people sign their names. But the ones who didn't — they were constructive in a way they wouldn't have been otherwise."

That's Layer 2. Knowing what Szymon needs. Designing the conditions that make honest feedback possible and having evidence-based conversations that actually move him forward.

What actually changed and what didn't for Lukasz

So I asked Lukasz the question I really wanted answered.

Before AI, I said, I know you were already a good manager.

He said with a modest smile. "I'd like to think so. When I applied myself fully, yes."

And after AI, what actually changed?

"The quality of a singular conversation? Not much. What changed is consistency. I have fewer weaker moments. If I had a hectic 2 weeks, I'd skip prep for one or two 1:1s. Now that doesn't happen. AI doesn't get tired. It does exactly the same thing every time."

I did not see that coming. And as someone who loves consistency — I should have. 😅

Not faster. Not smarter. More consistent.

Before AI, Lukasz spent the majority of his prep time on Layer 1; pulling notes, tracking patterns, and processing feedback. Now he spends that time on Layer 2. Thinking about the person. Reflecting on what they need. Deciding what the data actually means.

#nosurprises – from managers

For the employee on the receiving end of a manager like Lukasz, that's exactly what it feels like. Reviews arrive without surprises. Recognition lands when it's earned. Just consistent, evidence-based conversations that make growth feel inevitable, and performance feel fair.

AI doesn't make a good manager better in the moment. It makes them consistently good across every moment, and that is what changes the experience for the people they manage.

Step up from HR Leader to Business Leader

Ready to influence strategically, drive business impact and make HR indispensable?
Here are 3 ways I can help: