-1 to 1: Are VCs moving earlier?

In a recent conversation with a seed VC and an LP evaluating early-stage strategies, a common theme emerged: VCs are going earlier - pre-signal, pre-idea. The trend isn’t just seed-stage aggression - it’s a structural response to AI flattening the sourcing landscape.

Historically, top-tier VCs had a proprietary edge in sourcing: inbound flow, elite networks, pattern recognition. Today, tools like Harmonic compress that advantage. AI enables any fund to surface high-signal founders the moment they hit traction. In response, institutional VCs are moving upstream - toward the “-1 to 1” phase - where the competitive moat is founder access, not founder visibility.

Greylock’s $1B 17th fund is emblematic. Alongside the fund, the firm launched Edge, a program designed specifically for pre-idea and pre-seed founders. It’s not an accelerator - it’s a founder-capture play: custom support, early customers, recruiter access, and capital, all before company formation.

In the UK, Concept Ventures raised a £50M pre-seed fund with a thesis focused on founder-market fit well before product. They’ve noted larger funds dipping into seed rounds, pushing specialist funds to capture relationships earlier.

Venture builders like Cambridge Future Tech are formalizing this trend in deep tech. By identifying latent IP in academia and spinning it into startups with embedded teams, they’re not waiting for companies - they’re co-founding them. Similarly, Creator Fund has deployed student investors across 32 European campuses to scout PhDs and researchers at the ideation stage, offering up to €800k per investment.

Carta reports that in 2024, U.S. pre-seed startups raised ~$4B across over 25,000 convertible instruments. Nearly 44% of Q4 pre-seed deals were under $250k - a jump from 30% a year earlier - indicating an uptick in pre-traction, pre-team capital deployment.

PitchBook’s 2023 valuation report shows that while late-stage and Series A valuations declined significantly post-2021, pre-seed and seed valuations remained resilient. Why? An oversupply of small funds chasing early-stage deals. The implication: investor competition is fiercest where signal is weakest.

NFX argues AI has made sourcing 5x more efficient for newer or less-networked investors. As public signals proliferate and sourcing becomes democratized, edge returns shift upstream - to judgment, founder relationships, and access at inception.

If AI has made sourcing easier, it’s made true access harder. The best founders now meet dozens of VCs as soon as they signal. To win, VCs must be there before the signal - either by backing individuals directly, embedding in talent ecosystems, or building programmatic ways to manufacture deal flow.

From founder-in-residence models to embedded university scouts, the strategy is converging: intercept talent before company formation. That’s where proprietary deal flow still exists.

In this new paradigm, conviction and value-add replace pattern recognition. The best early-stage firms aren’t chasing deals - they’re creating them.