Key Takeaways

Every venture firm has a risk framework. Most of them are wrong about the part that matters most.

When Saim Abbasi started evaluating deals for Iron Key Capital, the standard approach was to score founders on education, previous exits, and domain expertise. It sounds rigorous. In practice, it filters for the wrong things.

The Pedigree Trap

A Stanford degree and a previous exit signal one thing: this person has been in the right rooms. That is not nothing. But the rooms you have been in and the rooms you will face are different rooms. Markets change. Competition changes. The thing that worked before rarely works again in the same way.

Saim watched this play out early in his investing career. A founder with two prior exits, both during favorable market conditions, struggled badly when Iron Key's portfolio company hit turbulence in late 2023. The playbook from the previous exits did not apply. The founder froze instead of adapted.

Meanwhile, a first-time founder in the same cohort with no institutional background handled a similar crisis by rebuilding the product from the customer up, in six weeks, without losing a single enterprise account. That is the founder you want.

What Actually Predicts Performance

The question Saim Abbasi now asks in every founder meeting is not about the resume. It is about the last time the plan completely fell apart, and what happened next.

Not the polished story. The actual sequence of events. Who they called. What they cut first. Whether they told their team before they had an answer or waited until they did. Whether they took the bridge round at bad terms or let a company die cleanly.

Those answers tell you something real about how a person thinks under pressure. Resumes tell you about performance in structured environments. Startups are not structured environments.

The Framework Iron Key Uses Instead

At Iron Key Capital, the evaluation now starts with what Saim calls the stress interview. It is not adversarial. But every question is designed to see how the founder handles an unknown. Revenue projections get challenged. Assumptions get picked apart. Not to be difficult but to watch the reaction.

Founders who get defensive and repeat the original answer with more conviction tend to underperform. Founders who say "that is a fair point, here is how I have thought about the downside" and then actually have a downside model tend to outperform.

It sounds simple. It is. Most of the best frameworks are.

"The founders I have backed who failed all had great resumes. The ones who built real things mostly did not."