Reach Capital's Best Bets Started as the Ones Its Own Team Argued About
Venture Capital·October 21, 2026

Reach Capital has built its reputation on backing companies that can look odd on first read. In a conversation with Crunchbase News, partner James Kim said the firm's education background still shapes how it searches for deals, and that some of its strongest investments were once considered "weird, and maybe uninvestable" by the people who would later champion them.
The firm's approach to AI detection is a clear example. Kim described GPTZero, the company known for tools that identify AI-generated writing, as a deal that initially divided Reach's own team. Disagreement inside an investment group is common, but Kim's account suggests the debate was not a sign of a weak opportunity. For a firm whose thesis depends on seeing value before the market agrees, a split room can be a useful signal in its own right, especially when the question is whether a product that sits at the edge of a new category will find a durable market.
Kim also explained how the firm thinks about getting its money back. Reach has relied on ownership stakes that let it stay invested as companies grow, and on secondary transactions, in which existing shares are sold to other buyers before an acquisition or public listing. Those routes can return capital on a different timeline than a traditional exit, which matters for an early-stage investor whose results may take years to show up in cash.
The discussion offers a window into how a smaller venture firm weighs conviction against consensus. For founders, the takeaway is that an investor's willingness to back an unusual idea may depend on its own history and on how comfortable it is with internal disagreement. For other investors, the conversation is a reminder that the most promising deals are not always the easiest to explain at first.
Reporting based on an external source.