The Moment The Data Stops Cooperating
At some point in almost every drug discovery program, you reach a moment where the data in front of you refuses to match the narrative you pitched. Maybe a lead candidate that looked pristine in primary cell assays falls apart once you test it in a more complex system. Maybe an engine target that appeared entirely selective starts showing off-pathway effects that no literature search flagged.
When this happens, every team faces the same defining fork in the road: Do you keep pushing the program forward to justify the time and capital already spent, or do you stop and listen to what the biology is actually telling you?
I have sat on both sides of that table. I have lived through the pain of shelving work I cared about, and I have also watched teams talk themselves into a fragile consensus simply because nobody wanted to be the one to pull the plug.
The Gravity of Sunk Cost
A program that has consumed a year of high-intensity work, significant funding and dozens of FTE hours generates its own gravitational pull. People start arguing for a molecule the way they would argue for a colleague. Nobody wants to be the reason six months of effort gets shelved.
I understand the instinct, it is human. But protecting a program because of what it already cost is not the same as protecting the science. Those are two different questions, and good teams learn to keep them separate.
When I evaluate this crossroads, I use a simple heuristic: What would this data tell me if I had walked into the room this morning?. If an objective set of eyes would call it a no-go, then twelve months of prior effort is not a reason to persist. It is a reason to be more honest, not less.
What A Disciplined Good No-Go Decision Actually Looks Like
A no-go call is not a failure of the science. It is often the clearest, most definitive signal the biology can offer. When a lead modality loses its pharmacodynamic window or triggers an unmanageable immunogenic signature, the system isn’t being uncooperative.. That is the biology handing you a very specific and precise map of boundary conditions- telling you what can be engineered, what cannot , and what needs to be set aside. Disciplined teams treat negative findings with the exact same rigor as positive breakthroughs:
- No repetitive reruns: Resist the urge to rerun an assay five times in search of an anomalous green bar.
- Post-mortem documentation: Document the failure mechanism thoroughly so future platform teams don’t rerun the same experiment under a different name two years later.
- Decouple ego from outcome: Predefine the kill-criteria before running the experiment.
The best R&D groups establish their Go/No-Go gates when the protocol is drafted—long before anyone’s personal identity is wrapped around the result. When the data arrives, you aren’t debating whether to stop amid emotional exhaustion; you are simply measuring the readout against a threshold the consortium already agreed on.
The Real Price Tag: Opportunity Cost
There is a fundamental truth in drug discovery that isn’t said out loud often enough: an early no-go is cheap. A late no-go is catastrophic.
Every month a non-viable asset remains in live support, it silently consumes high-throughput automation capacity, reagent budgets, and–most critically—the intellectual bandwidth of scientists who could be driving a program with a viable clinical path forward. That loss rarely shows up as a single, dramatic write-down; it shows up as an imperceptible drag across the rest of the portfolio..
True capital stewardship is never just about preserving the budget allocated to the current program. It is about clearing the runway for the next breakthrough waiting in the wings.
Reframing the Decision
Calling a stop is not the same as admitting defeat. It is a working decision, made with the best data you have at the time, and it can be revisited if new information shows up.
I would rather have a team that makes a defensible no-go call early and moves on to something more promising, than a team that keeps a weak program alive because stopping feels like losing. The second path feels safer in the short term. It almost never is, once you look at what else could have been done with that time.
Disciplined science requires knowing when to walk away from a hypothesis that has done its job. Deciding to stop is not a lesser outcome than nominating a clinical lead—it is the exact same standard of scientific integrity, applied at the moment it matters most.