In a recent interview at TechCrunch Disrupt 2025, Jennifer Neundorfer, co-founder of January Ventures, offered a sharp wake-up call to AI-startup founders: in the current environment where everyone is chasing “AI,” it’s not enough to build just a little better. 

What catches her attention is a startup that uses AI to create a new experience or workflow rather than merely improving the status quo. Neundorfer emphasised three key traits she’s looking for:

1. Founders who articulate why their idea is truly different from “dozens of startups doing that same thing.”
2. Teams that understand what their customer actually wants, not just what AI can do.
3. A readiness for a potential downturn: she expects a market correction in AI funding, and believes the winners will be those building “category-defining” companies, not chasing hype.

Search More This News? Click Here 🔍

Analysis & Opinion

Neundorfer’s commentary taps into a broader shift in how VCs and founders are behaving in 2025’s AI-boom climate, something that matters for U.S. startups and their ecosystems.

1. From “better” to “new”
Her point that incremental improvements (even “10× better”) may not be enough sparks a necessary caution. In many sectors enterprise software, consumer apps, health tech, there are already countless AI plays promising marginal uplift: faster workflows, better metrics, slight automation. But when the space is flooded, differentiation becomes difficult and capital scarce. By saying she looks for “new behavior” or “new workflow,” Neundorfer is essentially requiring founders to ask: does this idea rewrite how something is done, rather than just speed up how something is done? That criterion is harder to satisfy, but in a crowded field, it may be the only defensible position.

2. Customer first vs technology first
The emphasis on “what the customer wants versus just building what is possible” hits a recurring tension: too many founders start with “we can build with AI,” rather than “customers care enough to adopt this.” From a U.S. investor perspective, that matters because many AI startups die not from lack of technology, but from lack of traction or differentiation. Neundorfer’s framing suggests founders need to ground their story in concrete insights about behavior change, adoption, and value not just “we’ll apply a large language model because we can.”

3. Timing, liquidity and market signals
In her remarks about a possible correction and “winners building category-defining companies,” Neundorfer gestures to the cycle risk in venture. 2024-25 have seen a huge surge in AI interest, but with that comes a glut of companies, high valuations, and rising expectations. The implication: founders who enter now must be more rigorous, more future-looking, more able to ride the next wave (not just the current one). For U.S. founders, this means paying attention to runway, to defensibility, to whether the category they target is truly open.

4. Relevance beyond the U.S.
While Neundorfer speaks from a U.S./Silicon Valley-adjacent context, the logic holds globally. For any startup building AI in crowded markets (including emerging ecosystems), this means: don’t assume “just throwing AI at the problem” will attract investors. Local founders must also consider how their context gives them a unique behavior insight or workflow shift. For example, local regulatory, cultural or infrastructure conditions may create opportunities to leapfrog incumbents, but only if the founder frames it in a way an investor can understand.

5. Implications for founders and ecosystem players
Founders should revisit their pitch decks: is the claim “10x faster” or “new behavior”? Can you clearly articulate why your product will be adopted, who will adopt it, and how they’ll change behaviour?

Investors appear to be tightening their filters: in sectors crowded with entrants, they’re looking for clear differentiation, defensible market position, and realistic funding models.

Ecosystem enablers (accelerators, incubators, policymakers) may need to recalibrate their support models: helping startups define unique workflows and business models may become more valuable than simply “apply AI to data.”

Final Summary

In short, Jennifer Neundorfer’s advice offers a timely reset for founders operating in the saturated arena of AI startups. Rather than chasing incremental gains, she advocates for bold differentiation building new workflows or behaviors, deeply understanding customer needs, and preparing for a market that may tighten. For U.S. founders, this means intention, clarity, and focus. 

And for the broader ecosystem, it signals that the next wave in AI won’t just be about bigger models or more data it will be about redefining how things work. If you’re founding now, ask yourself: what behavior will change because of my product? If the answer isn’t compelling, you may be fighting the tide.