‘Weird And Maybe Uninvestable’: How Reach Capital Spots Its Venture Winners

The Beauty of the Bizarre

In the venture world, consensus is usually the kiss of death. If everyone in a room agrees a startup is a slam dunk, the valuation is likely already pushed to the moon and the edge is gone. I recently dug into the investment philosophy of James Kim at Reach Capital, and it confirms something I have felt for a long time: the real winners are often the ones that feel a bit weird at the start.

Reach Capital started with a narrow focus on education technology. For years, edtech was the stepchild of Silicon Valley—slow sales cycles, bureaucratic hurdles, and limited exit opportunities. But staying in the trenches of a difficult sector gives you a different perspective on what makes a business resilient. Now, as they expand into broader AI and healthcare, they are using that same skeptical, builder-focused lens to find companies that most generalist funds are too scared to touch.

The Polarization Metric

One of the most interesting takeaways from Kim’s recent insights is the idea that internal disagreement is a buy signal. At Reach, some of their most successful exits and high-performers were companies that initially divided the partnership. Half the room thought it was the future; the other half thought it was uninvestable. This is where the alpha lives.

For founders, this is an important lesson. If you are pitching a room and getting polite nods from everyone, you might not be disruptive enough. You want to provoke a reaction. You want to be the reason two partners argue in the hallway after you leave. That friction usually means you are challenging a fundamental assumption about how a market works. If your business model doesn't make a traditional VCs stomach turn just a little bit, you might just be building a feature, not a company.

The GPTZero Case Study

Take their investment in GPTZero. When AI writing first exploded, the immediate reaction from the hype-cycle crowd was that detection was a losing game—a temporary cat-and-mouse chase that OpenAI would eventually solve themselves. But Reach saw it differently. They saw a fundamental need for trust and transparency that went beyond just catching kids cheating on essays.

They looked at the builder behind it, Edward Tian, and realized that the value wasn't just in the current algorithm, but in the positioning. While everyone else was rushing to build the next wrapper around GPT-4, GPTZero was building the infrastructure for verification. It felt niche, it felt like a tool that could be Sherlocked by Big Tech, and yet it has become a cornerstone of the AI integrity stack. That is the reward for betting on a 'weird' use case before the rest of the market catches up.

Beyond the Hype: The Real Cost of AI

We are currently living through a period where every pitch deck has 'AI' in the first three slides. Kim and the Reach team are looking past the buzzwords. They are asking the same questions I ask: How do you actually make money when your compute costs are scaling as fast as your user base? How do you defend a moat when the underlying model is owned by a third party?

The shift from edtech into healthcare is a natural progression here. Both sectors are heavily regulated, rely on high-stakes human outcomes, and are notoriously difficult for outsiders to navigate. If you can build an AI tool that actually improves patient outcomes or reduces administrative burnout without breaking HIPAA, you have a real business. But getting there requires a founder who understands the 'boring' parts of the industry—the plumbing, the billing codes, the compliance.

Liquidity and the Long Game

One of the most candid parts of Kim’s strategy revolves around how they return capital. In a market where IPOs have dried up and M&A is under heavy regulatory scrutiny, venture firms have to be smarter about secondary transactions. Reach isn't just sitting on their hands waiting for a billion-dollar exit. They are looking at ways to lock in gains and manage ownership stakes strategically.

This matters for builders because it changes the dynamic of your cap table. You want investors who understand that the path to a win isn't always a straight line to a Nasdaq listing. You want partners who are disciplined about their own fund economics, because that discipline trickles down to the advice they give you. A VC who is desperate for a home run to save their fund will push you to take reckless risks. A VC who managed their liquidity well can afford to be patient while you find your product-market fit.

Takeaway for Founders

If you are building something right now that feels 'weird' or is getting rejected by generalist funds for being too niche or too difficult, don't pivot just yet. The most resilient businesses are often the ones that were initially labeled uninvestable. Focus on the hard problems in unglamorous sectors like education and healthcare. If you can survive the skepticism, you’ll find that the competition is much thinner on the other side.

The goal isn't to be liked by every VC; the goal is to be the most important company in a specific, misunderstood room.

Stop trying to fit the mold of what a 'venture-scale' company looks like according to a 2021 playbook. The 2024 playbook is about efficiency, sector expertise, and the courage to stay weird.

Read the original at Crunchbase News →