Nate Singer

Announcing HERON INTELLIGENCE When I first came across Blue Heron many years ago, I recognized a rare formula for finding the "ground truth" through sophisticated, investigative conversations.   It was the exact "human access" edge I had been looking for as an investor.   Since then, my focus has been on bringing in depth qualitative research into the modern era - investing in team, tech, and developing an industry-leading set of capabilities. This week, those pieces officially come together. I’m thrilled to announce that Blue Heron is now Heron Intelligence. We are solving the problem of collecting, analyzing, and consuming the highest quality human-centric intelligence at scale. By merging 20 years of investigative rigor (and a massive archive of 75,000+ proprietary interviews) with a new AI-driven intelligence layer, we are establishing a best in class capability suite for qualitative management and business insights.   To our clients and partners: The mission hasn’t changed, but the tools have been substantially upgraded. This means more signals, less noise, and the same investigative intensity you’ve always relied on.   To the Heron team: You are the engine behind this legacy. I couldn’t be prouder to lead this team as we bring data, custom analysis, and leading AI tech to the world’s top asset managers.   The future of qualitative intelligence is here. prnewswire.com/news-releases/…

Today is the last day of the quarter, which means earnings season is in a few weeks.   As an analyst / PM, this was always the most exciting - and anxiety-ridden - time of the year. A brief release, scripted Q&A, then lots of volatility with not much time to get your work done.   How can AI help if the data doesn’t even exist? Serious analysts will be super excited by Heron’s upcoming release - and get a real advantage during that risky post-earnings information vacuum.   Stay tuned!

If you had read the Heron transcripts in January 2015, you wouldn't have been surprised at the stock chart today. That's the value of diligence that goes beyond the model.

The window between when you can see the problem qualitatively and when the market prices it quantitatively is where the alpha lives. And that window is often years. Sometimes a decade. A 2015 interview isn't old. It's a blueprint.

How can a 2015 management diligence report still be useful in 2026? Because people don't change. Culture doesn't change. And the way aggressive, desperate management behaves - the shortcuts, the red flags - those are stable. And so, to those who know, the eventual outcome isn't a surprise. It can just take a while to arrive.🧵

@mbateman @jliemandt It is just crazy high number Lalilo isn’t super AI or in-house. Same with most of the other apps. Is that your Claude development tokens bill divided by the number of active students? Implies that Oslo’s daily token usage is $500 — doesn’t make sense

@NateSinger_BH @jliemandt Yes

This will be true of schools too. Alpha School spends over $10k per student in AI tokens.

@JaredSleeper Are you saying there is more signal in 150-200 expert calls? Doesn’t really work that way. Reminds me of a joke… What is worse than finding a worm in your apple? Finding half a worm. What is worse than finding half a worm?

In research orgs, by far the biggest time-waster I see is “gathering n” via manual hourlong expert calls that often just reduce down to a few datapoints. I’m sorry but even 15-20 calls is unlikely to be statistically significant on anything that counts as a debate.

@Accelr8_Dan @27XVII For sure, we do - most of Heron’s services revenue is for top tier PE. I’ve kind of got my hands full with this private company right now. With that said, public markets are 100x more actionable / scalable. For real - how can you short bad PRIVATE company management teams?

@NateSinger_BH @27XVII Why not just exploit the even greater information asymmetry in private markets?

Everyone talks about information edge in investing. Here's where the edge actually lives: knowing the CFO, two division presidents, and the head of product all left in the same 6 months. The financials show this eventually. But you want to know first. How?

If you’d like to see a third party backtest illustrating the strength of this “market inefficiency”, across the largest management diligence interview dataset in the world, reply BACKTEST

What are you seeing as best practices in evaluating management?