Most venture diligence checklists were built for software. Growth rate, retention curves, CAC-to-LTV, competitive moat, team pedigree — the same seven or eight boxes, checked in the same order, for every company in the pipeline. It works reasonably well until the product is making clinical decisions about a human being in crisis. Then the checklist that caught every other risk in the deal goes quiet exactly where it matters most.
The Gap Standard Diligence Doesn’t Cover
A slick interface, a strong NPS score, and a compelling founder story can sit on top of a clinical model that has never been validated against anything. Nothing in a standard SaaS diligence process is built to catch that, because nothing in a standard SaaS diligence process is looking for it. The result is a portfolio company that looks investable on every metric a growth-stage associate knows how to read, and uninvestable on the one a licensed clinician would flag in the first meeting.
A slick interface and a strong NPS score can sit on top of a clinical model that’s never been validated against anything.
I run three questions on every mental health deal before it goes further. None of them show up on a standard term sheet checklist, and all three have ended conversations that everything else about the deal made look easy.
Question One — Is the Intervention Actually Grounded in Something?
Is the product’s core intervention rooted in an evidence-based modality — CBT, DBT, ACT, a validated protocol with a research base behind it — or is it a well-designed interface wrapped around technique the team picked up from a training weekend? The two can look identical in a demo. They do not look identical eighteen months later, when outcomes data either backs up the product or it doesn’t.
Question Two — What Happens When Someone Is in Crisis?
What’s the protocol when a user shows signs of suicidality or psychosis, and is that protocol legally compliant across every state the product operates in? This is the question founders are least prepared for — not because they don’t care, but because building a crisis protocol that holds up across dozens of different state licensing regimes is genuinely hard, unglamorous work that rarely makes it into a pitch deck.
Question Three — Who Actually Owns the Outcomes Data?
Who owns outcomes measurement, and what does the data actually show? Not “we track engagement” — real clinical outcomes, tracked consistently, owned by someone with the clinical background to know whether the numbers mean what the team thinks they mean.
Why This Belongs in Diligence, Not After the Close
Every one of these questions is answerable before a term sheet is signed. They just require someone in the room who knows what a real answer sounds like versus a well-rehearsed one. That’s the entire job — not slowing down good deals, but making sure the deals that look good on a SaaS scorecard are actually good on a clinical one too.
I ask these three questions on every deal I review. If you want a second clinical read on something already in your pipeline, that’s a conversation worth having before the close, not after.