Data is not strategy.

There's a belief that more is safer. More studies, more endpoints, more confirmation — each addition feels responsible, each dataset feels like progress toward completeness. So teams respond to uncertainty by accumulating data.

But it backfires more often than anyone admits.

Here's how it happens:

A repeat-dose study turns up mild, non-progressive liver enzyme elevations at the high dose — no histopathology, exposures well above your margins, not adverse. On its own, that result is interpretable and defensible. You could write the rationale today.

But the team never pre-defined how biochemical changes get weighed against exposure margins. So, to "be safe," they add a follow-up study. The enzyme changes recur — in a slightly different pattern, at slightly different incidence. Now there are two datasets that don't cleanly agree, and a finding that was a footnote has become a discussion. The extra study didn't add clarity. It added ambiguity you now have to explain.

That's the trap: data only reduces uncertainty when it's generated inside a clear interpretive framework. Without one, more data just expands the surface area of things you now have to justify and a reviewer can question.

And it's usually done in the name of a myth — that regulators expect everything. Every study, every endpoint, every angle covered. But regulators don't expect completeness, they expect intentionality. A program with a reason for every study beats one with a study for every fear.

I wrote Data Is Not Strategy to challenge the assumption that volume equals rigor. The heart of the book teaches why interpretation, not volume, drives nonclinical decisions, and how to build programs that answer the right questions instead of all of them.