Your data package is your pitch: what partners and investors check first

I spend most of my week on the CRO side of preclinical programs: scoping the studies, writing the quotes and SOWs, and seeing what the data package looks like when it's done. The same package is what a pharma search and evaluation team or a venture investor opens after the first meeting goes well. They read it with one question: does this hold up?

Here are the seven checks I see matter most, and what a ready answer looks like for each.

1. Does the efficacy reproduce?

One striking result in one experiment is a signal, not a finding. Reviewers look for independent repeats, adequate group sizes, randomization and blinded scoring, and ideally a second model or a second endpoint pointing the same way.

Ready answer: "We've seen it in three independent studies, in two models, with blinded readouts. Here are the group sizes and the statistics."

2. Is the model relevant to patients?

Every disease model is a compromise. Diligence asks whether yours reflects the patient population you're targeting and whether its endpoints map to something a clinical trial could measure.

Ready answer: a short paragraph on why this model, what it misses, and which translational biomarker bridges it to the clinic.

3. Is exposure linked to effect?

This is the gap I see most often. Efficacy studies run without PK sampling leave a reviewer guessing whether the drug reached the target at all, and at what level. Without exposure data, a dose is just a number.

Ready answer: plasma (and ideally tissue) exposure from the efficacy studies, plotted against the effect, with target engagement or a PD marker if you have one. Adding satellite PK animals to an efficacy study is usually a modest line on a CRO quote.

4. Is there a dose response?

A single active dose leaves open whether you're at the top of the curve, the bottom, or somewhere unhelpful. A dose response tells a partner you understand the pharmacology and gives the clinical team a starting point.

5. What's the early safety margin?

Nobody expects GLP toxicology at proof of concept. They do expect an early read: a dose-range finding or tolerability study with exposure data, compared against the exposure you need for efficacy. That ratio is the first estimate of your therapeutic window, and reviewers will calculate it whether you do or not.

6. Can you make it again?

Diligence teams ask which lot of test article went into which study, how it was characterized, and whether it can be made at scale. An efficacy result from an uncharacterized batch is hard to build on.

Ready answer: a one-page CMC summary with identity, purity and stability data, the lot used in each study, and the manufacturing route to clinical supply.

7. What's GLP and what isn't?

Most early data is non-GLP, and that's fine. What costs credibility is presenting exploratory data as if it were definitive. Label each study clearly, and say what the IND-enabling package will add.

How to present it

Put one table in the confidential deck and the data room: every study, the model, group sizes, doses, the result, GLP status and whether the final report is done. It takes an afternoon to build. It answers a dozen diligence emails before they're sent, and it shows a team that knows its own data.

If the table shows gaps, that's useful too. Most of them can be closed with a focused study, and it's better to close them before outreach than to explain them in the first diligence call.