Happy Thursday, and welcome to BeyondTVPI.

Later this month, I’ll be visiting one of the large Danish Pension Funds to do a deep dive on their investment strategy and what could be improved or systematized.

It made me revisit the topic of the first GP meeting and how we did it. Naturally, we had written goals and processes for it.

A first GP meeting is 60 minutes. The GP arrives with a deck and a plan. They know what they want to cover, in what order, and which slides to linger on. They have answers ready for the hard questions.

But, how does an LP arrive at a meeting?

For me, the answer was always the same: I needed to be able to rate this GP across five areas - investment strategy, value creation, organization, results, and terms - and feel confident enough in each score to defend it.

That became my defined outcome. Everything I did in the meeting followed from it. And everyone across the organization did the same.

Over time, this paid dividends as valuable discipline and infrastructure, that ensured we evaluated every manager the same - no matter who participated in the meeting.

Before I asked anything about the GP, I talked about us.

Three to five minutes. Who we were, what we invested in, a few of our prior commitments - named, specific.

This wasn't courtesy. It was positioning.

GPs are not just evaluating your cheque. They're evaluating whether you're a professional, informed LP. Whether you understand their strategy. Whether you're a tourist or not.

In a world where access to the best managers is genuinely scarce, that three-to-five minutes is probably the most underused move in a first meeting.

Results and terms I usually had from the deck already, or I'd ask briefly if not. One thing I always made sure to get, regardless: their expected fundraising timeline. When are they next in market? What's the target size? A GP who says "probably 2027, targeting $1.5bn" and comes back in 2026 at $2bn is telling you something - but only if you wrote down what they said the first time.

But that's not where the meeting lives.

The real work was investment strategy, value creation, and organization. And across all three, I was testing for one thing: a red thread.

Not a strategy covering consumer, industrials, and services with some turnaround here and some growth there. A clear, specific identity: this is who we are, this is what we do, this is why we're better at it than anyone else.

The model itself didn't matter much. Buy-and-build, CEO-first, operational turnaround - all of these can work. What I was looking for was whether the GP could articulate their edge with precision and consistency.

The commitment that almost never was

I remember one meeting where we were about 20 minutes in, heading into case studies. We'd covered strategy, value creation, organization. Their returns were good - I already had those. But I stopped and realized I couldn't do it. I couldn't rate them. I couldn't articulate who they were or what made them different.

So we didn't go to the case studies yet. We spent another 10 minutes on their process - how they actually operated inside a company, who drove what, what their playbook looked like in practice.

Then we did the case studies. And suddenly everything clicked. What had seemed like a competent but unremarkable pitch became coherent. The ratings fell into place.

That turned out to be one of my best commitments.

Seven years on, it's clear first quartile on TVPI, IRR, and DPI - and among the relatively few post-2019 vintages already showing real DPI.

The 10-minute detour was the meeting. The case studies just confirmed what I had already started to believe.

After every meeting, I'd set a rating - A to E - across each of the five areas, then roll it up to one overall score.

It wasn't scientific. But it forced me to confront each dimension explicitly rather than letting the parts I liked colour everything else. And over time, the discipline of scoring every meeting meant my judgement got better - the rating after a first meeting was rarely dramatically off from where we ended up after full diligence.

That's the payoff of having a defined outcome: it trains your judgement across hundreds of meetings, not just one.

Most experienced LPs already do some version of this, informally. After enough meetings, you develop a feel - and some are genuinely good at reading a GP in a room. The issue isn't the assessment. It's that it stays in their head.

A rating that lives in a notebook, or worse, in someone's memory, depreciates. Every personnel change, every handover, every re-engagement with a GP years later - the context degrades. What should be institutional memory becomes personal memory. And personal memory is not infrastructure.

The question I'd sit with: could any member of your IC reconstruct a ten-year GP relationship history - independently, from your records, within minutes? If the answer is uncertain, the issue isn't the meetings. It's what happens after them.

I know some LPs will disagree with this.

There's a view that LP judgement is instinctive. That judgement is earned through experience. That formalising it risks oversimplifying something nuanced.

I understand that instinct.

I still believe it's wrong.

"More art than science" too often becomes an excuse not to define standards. Not to write things down. Not to force clarity.

The goal isn't perfection. It's discipline.

Write something. Use it. Watch where it breaks. Improve it.

Over time, that simple act compounds. It sharpens questions. It sharpens listening. It sharpens pattern recognition.

And it doesn't stop at first meetings. The same principle applies to reference calls, IC memos, re-underwriting managers, even portfolio reviews. Clear outcomes and lightweight SOPs elevate judgement and ensure you actually turn the important stones - they don't constrain it.

If you want a practical breakdown of what to capture during and after a first GP meeting, I've written that up in more detail here.

Before your next first GP meeting, write down one sentence: "After this meeting, I need to be able to..."

For me, that sentence ended with: rate this GP across five areas and feel comfortable defending each score.

That's your defined outcome. Everything else - what you ask, what you listen for, when you slow down and ask for 10 more minutes - follows from it.

Last week, as we were pushing a new feature live, our tests failed.

My first reaction was: This is fantastic.

Not because something broke. Because the system caught it - exactly as it was designed to do. Honestly, I was more excited about that failing test than I've been about shipping some features.

I love good infrastructure. I love checks and balances. I love taking one step back, to take two steps forward.

This test will run forever, just like all the other tests we’ve set up. As we build increasingly complex software, this will pay increasing dividends over the years.

We build systems that test the systems. And when they work - when a test fails because it's supposed to - it's not frustrating.

It's exactly the point.