Happy Thursday and welcome. Let's dive in.

Last week a PE VP asked me a question I hadn't heard put quite that directly before.

"How do you actually do reference calls with former employees of a firm?"

My answer: If a random LP you'd never met reached out to you asking about your former employer, would you tell them the real story?

He laughed. He got it immediately.

That's the problem in one sentence.

Former employee references are structurally broken. Not because the people you're calling are dishonest — most of them are perfectly good people who left for perfectly reasonable reasons. But the incentives around what they'll actually say to a stranger are almost entirely pointed in one direction.

They probably don't know you. You found them on LinkedIn, or from a list of names you got somewhere. They have no reason to trust you. They have plenty of reasons not to burn bridges. The PE industry is smaller than it looks, and the GP in question might be their next LP introduction, or their next co-invest partner.

So they give you a politician's answer.

"It was a good firm. Strong team. I learned a lot. The reason I left was really about wanting a different stage of company / sector focus / title opportunity."

They probably confirm, implicitly, that there was a reason they left. But they cover it in enough positive framing that the signal gets buried. You finish the call feeling like you did something productive.

You mostly didn't.

At best, this kind of reference does nothing. You hang up, the call is net neutral, and you move on.

The real risk is worse: it tips you toward confirmation bias.

Charlie Munger described it better than anyone: "The human mind is a lot like the human egg. When one sperm gets in, it shuts down so the next one can't get in." Or in other words, once an idea gets in (we like this GP), the mind starts defending it and resisting contradictory evidence.

Sidebar: I am fairly certain there was a heavy dose of confirmation bias detailed in this decision.

Here's why that matters. Most LPs enter a reference call at a moment when they're already leaning toward yes. The diligence is advanced. The deck has been reviewed. You liked the team in the meeting. You've already allocated mental energy to the investment.

When the former employee gives you a clean, positive-sounding answer, it doesn't just fail to challenge your thesis — it reinforces it. You heard nothing bad. The reference checks out. Your original conviction feels validated.

The call wasn't independent evidence. It was an echo of the conclusion you'd already reached.

The variable that changes everything is whether the relationship exists before the call.

When you call someone you genuinely know — a former colleague, someone you've stayed in touch with, someone who has reason to be candid with you — the conversation is different. The politician's answer evaporates. People will tell you real things when they trust the container.

Which means the value of former employee references isn't really in the reference itself. It's in the network you built before you ever needed the reference.

If you run into a PE VP at a conference, have a real conversation, and build an actual relationship — that person becomes a legitimate source two years later when their former employer shows up in your pipeline. You call them as a contact, not as a stranger. The conversation starts somewhere completely different.

If you cold-call them from a LinkedIn search with a generic intro, you're getting the politician.

None of this means you should stop talking to former employees. It means you should be honest about what you're actually getting.

A cold reference call with an unknown former employee tells you roughly this: they left, they're not publicly hostile, and the firm isn't a complete disaster. That's useful as a floor check, not as diligence.

The questions that actually help in these conversations aren't the ones that invite a politician's answer. Don't ask "what was it like working there?" Ask something narrow enough that a polished non-answer would be obviously absurd. "How did the team handle a deal that went wrong?" "What happened when a partner disagreed with the IC on a deal?" Specific, process-level questions are harder to polish than general ones.

And the best use of the call might not be the content at all. It's reading what doesn't get said. A former employee who gives you a 30-second answer and pivots quickly is telling you something. So is the one who pauses too long before answering, or who volunteers a very specific compliment about something very specific.

But none of that substitutes for having built the relationship first.

The PE VP who asked me that question is good at his job. He'd asked it because he'd noticed the problem himself — the calls that sound fine but feel hollow. That instinct is right.

You can close some of that gap by asking the process questions.

But, he gap between a reference call that confirms your bias and one that actually challenges is less driven about what you ask.

It's about who picks up the phone.

The opportunity cost of rest has gone up.

A free weekend used to feel like a gift. Now it feels like a decision. If Saturday has nothing on the calendar, some part of my brain immediately starts calculating what I could ship by Sunday night.

It’s been many years since I’ve felt like this.

When I was 16, I'd sit in history class mentally planning my Diablo II character build. Not because the class was boring. Because I was completely obsessed. I couldn't stop thinking about it even when I was supposed to be somewhere else.

That's exactly what building with AI feels like right now.

TV shows I'd have been fully absorbed in two years ago — I'm half-watching them, mind already drifting to a product decision or a feature I want to spec out. My wife notices. I notice. It's not burnout. It's the opposite of burnout. It’s excitement.

Before AI, a free weekend cost one weekend of human-speed work. Now it costs one weekend of AI-augmented work — a draft, a deck, a prototype. That's a fundamentally different calculation, and it makes every idle hour feel more costly.

As a 40-year old, I didn’t expect to reach this teenage-level of excitement again.