The Operating Partner
Posts
OP 319: Leverage, Not Lazy

OP 319: Leverage, Not Lazy

Why Human-Agent Teams Change the Scoreboard

Darren Herman
March 03, 2026

Happy Tuesday, OP Community-

This is a special week for me - my birthday is tomorrow, and on Thursday, I get to spend the afternoon on the Padel court with the Silicon Alley Sports community at our first in-person event of the year. You are probably sick of me talking about Silicon Alley Sports, but I’m 19 years proud of it… We play well and do well.

The play well is self-explanatory - ideally, I play well at Padel. The do-good part is that all monies left over from hosting our 2026 event roster are donated to an impact partner… which we’ll announce closer to our flagship event on May 18. If you are interested in learning more and joining us, even if you don’t play, you can do so here.

I’ve continued to have more fun building - I am in the process of building a virtual marketing office, which includes content generation, research, and other marketing and growth office needs. If someone in the OP community is also building something in this space, I’d love to chat. Feel free to reach out.

Watches: One of my latest pickups is a 2022 Zenith x Hodinkee El Primero Limited Edition w/Salmon Dial. I put it on a blue leather strap, and the watch is absolutely stunning. I also recently did an AMA on my watch Insta and received over 30 questions, a new record for me, and I answered about 20 of them in the time I had. If you are into AI and you do collect watches, here is a set of AI Prompts you can use as a watch collector - super helpful.

Here are a few links to get us started:

An Exclusive Trip Inside the NBA’s Combine for Front Office Talent (Marc Stein)
Bain & Co’s Global Private Equity Report (Bain & Co)
What’s the ROI on AI? (Harvard Business Review)
And In The Juxtapositions May Lie the Meaning (Rishad Tobaccawala)
Burger King’s Patty Is Really Listening (Shelly Palmer)

I think you will like my main feature below. It brings up the timely topic of AI and pairs it with my other favorite topic: the NBA. Enjoy it.

Thanks again for subscribing to the OP. I truly appreciate you taking the time to read it. As always, please do reach out with any questions or feedback.

Be well, do good.
Darren

The "Court Vision" of Human-Agent Teams

If you've been following the OP for the last few months, you know I'm obsessed with the transition from Point Solutions to Human-Agent Teams (HATs).

In the enterprise, we are moving past the "magic trick" phase where AI just writes a better email or crops a photo. We are entering a world where agents aren't just tools…they are teammates. They have agency. They have a "span of control." And as operators, our job is shifting from managing people to orchestrating a hybrid workforce that is half silicon.

But when I talk to portfolio CEOs about this, the magnitude of the shift often feels abstract. They ask, "Has this actually happened before?"

The answer is yes. But it didn't happen in a boardroom. It happened on the hardwood of the NBA.

From Descriptive to Agentic

Think back to the "Moneyball" era. For years, sports analytics was Descriptive. We used spreadsheets to tell us what happened. A player shot 38% from three; a team had a 102 defensive rating. In the business world, this was your standard BI dashboard. It was a rear-view mirror.

Then, about a decade ago, things got Predictive. We started using data to guess what might happen. This is where "Load Management" was born. Teams started using wearables to track mechanical stress and biometric data, predicting injury risk before a hamstring actually popped. Remember Moneyball?

Today, we are entering the Agentic era of basketball.

The "Leverage Score" and the Counterfactual Agent

The NBA recently introduced something called the Leverage Score. This isn't just a stat; it's a real-time counterfactual engine. For every single possession, an AI agent runs thousands of simulations in the background, asking, "What if?"

If Jalen Brunson (yes, had to mention him) makes a cross-court pass instead of taking the floater, the agent calculates how much the win probability shifts in that exact microsecond. It credits players for "high-leverage" actions that never show up in a box score, like a defensive rotation that forced a pass into a low-percentage zone.

In the enterprise, this is the Holy Grail of HATs. Imagine an agent in your sales org that doesn't just track closed deals, but runs simulations on your entire pipeline to tell you which "low-leverage" meetings are actually eating your margin, or which "quiet" customer success touchpoints are actually the ones preventing churn.

The New "Orchestration Layer"

In basketball, the "Manager" (the Coach) used to be the sole source of truth. They operated on "gut" and "eye test." Today, the best coaches are Orchestrators.

They have a "data bench" that functions exactly like the Human-Agent Teams I've been preaching about. Look at the San Antonio Spurs. They are reportedly using AI agents to handle full-season travel logistics-a nightmare of variables like recovery windows, hotel altitude, and flight durations. What used to take a human staff three weeks now takes 20 minutes.

The human staff didn't lose their jobs; they moved to the high ground. They stopped being data-entry clerks for Delta and started being "Experience Architects" for the players.

This is the exact "Org Design" shift I'm talking about for your marketing and growth teams. If your managers are still spending 40% of their time "orchestrating humans" through status meetings and manual approvals, you are playing 1990s basketball in a 2026 league.

The Three "Watch Outs" for the HAT Era

As we move toward this hybrid model, the lessons from the NBA front office apply directly to the C-suite:

Don't Automate Taste: An agent can tell you the statistically "correct" lineup to stop a pick-and-roll, but it can't feel the "vibe" in the locker room. In marketing, an agent can optimize a headline for a 0.2% CTR lift, but it can't tell you if that headline makes your brand look like a commodity. Humans own the Narrative.
The Span of Control Problem: If one coach can manage 15 players, how many "agents" can a VP of Growth manage? We are finding that cognitive overload occurs more quickly when you're supervising "perfect" machines that never sleep. You need new accountability models.
Proprietary Context is the Moat: Every NBA team has access to the same optical tracking data. The "Alpha" comes from how they weave that data into their specific culture and playbook. In your business, the LLM is a commodity. Your proprietary data and your team's orchestration skill are the only things that aren't.

We are in a period of rapid experimentation. Whether you're trying to win a championship or hit a $14B fund target, the goal is to leverage, not be lazy.

We’re in the early quarter of the game here, but some real things are happening, and those who figure it out will have an advantage.