Use case · decision gate
Decide it once, and record why.
A go/no-go gate should end with more than a decision. It should end with the assumptions that decision rests on, the risks that would reverse it, and a signed record — so that when the evidence moves, someone knows to look again.
A gate settled on evidence — a no-go is a result too.
The situation
The meeting decides. Nobody records why.
The gate happens, a decision is made, and the reasoning lives in a deck and three people’s memory. A year later, when the assumption behind it has quietly stopped being true, there is nothing to re-open.
- Pillars behind the recommendation
- 6
- Analyses, each cited
- 30+
- E-signed decision, flagged when evidence moves
- Part 11
What happens today
The gate rewards the most confident voice.
When the evidence is uneven and the deadline is real, decisions get made on the strength of advocacy rather than on the strength of the case.
Advocacy substitutes for evidence
The person who has lived with the program longest presents it. Their conviction is genuine and it is also the least measurable input in the room.
The assumptions are never written down
A go decision rests on a handful of load-bearing assumptions — enrolment rate, comparator, price. If they are not listed, they cannot be monitored.
No record survives the meeting
The deck is archived; the reasoning is not. When the question comes back, it is re-argued from the start with a different cast.
A no-go is treated as a failure
Because stopping is culturally expensive, marginal programs continue by default. A well-evidenced no-go is the cheapest good outcome available, and the hardest to reach.
What you do instead
A recommendation you can interrogate.
A structured go/no-go recommendation
Everything from the six pillars rolled into one recommendation, with the interpretation written out — what it means, and what it does not.
A directional probability, with its arithmetic shown
The base rate it started from and every adjustment applied to it, listed separately — so the debate is about a specific adjustment, not about a score.
The risks that would change the answer
An operational-risk view, adverse-event signal and sponsor delay signals from public filings — the things that turn a go into a no-go later.
Load-bearing assumptions, listed
The assumptions the recommendation rests on are stated explicitly, so they can be monitored rather than rediscovered.
A decision memo, drafted
The CMO / VP R&D workspace drafts the go/no-go memo from the cited assessment — for your team to edit and own, not to rubber-stamp.
A signed record that re-opens itself
The decision is signed under 21 CFR Part 11. If the underlying evidence later changes, the signed decision is flagged for re-review automatically.
What runs, and who drafts
The scope for this job
A gate is the one moment all six pillars have to be on the table at once — the verdict is only as good as the five assessments feeding it.
Pillars involved
How it runs
Assemble, decide, and sign.
Put the program in
A protocol, a synopsis, or a molecule and an indication. The assessment runs across all six pillars against public evidence.
Interrogate the recommendation
Open any figure to its source. Challenge an adjustment to the probability. Read the comparable trials that failed and the risks flagged against this one.
Sign the decision
Your team approves the memo with a Part 11 e-signature, on a hash-chained audit trail. When the evidence behind it moves, the decision is flagged for re-review.
What lands on your desk
A decision record, not just a decision.
The output of a gate should be re-readable by someone who was not in the room. This one is.
What lands on your desk
Go/No-Go memo + signed decision record
- The recommendation, with an interpretation section stating what it does and does not support.
- A directional probability of success with its base rate and each adjustment shown separately.
- The pillar-by-pillar findings that drove it, each cited to a public record.
- The load-bearing assumptions, listed so they can be monitored.
- An operational-risk view, adverse-event signal and delay signals from public filings.
- A 21 CFR Part 11 e-signature and hash-chained audit trail, with automatic re-review flagging when the evidence changes.
Honest limits
Where this stops
What this does not do for you. Naming it here is cheaper for both of us than finding it in a procurement review.
The recommendation is decision-support. The decision is yours, and the signature records that it was yours.
The probability of success is directional — a structured argument from base rates and trial-specific evidence, not a forecast.
It reads public evidence. Your internal data, your strategic context and your portfolio constraints belong in the room alongside it.
Re-review flagging tells you the evidence moved. Deciding whether the move matters is a judgment the system does not make for you.
It is not a medical device and does not make clinical or regulatory determinations.
Questions
The ones you would ask first
What exactly gets signed?
The decision memo, bound to the evidence snapshot it was built from, under a 21 CFR Part 11 electronic signature on a hash-chained audit trail — so the record shows who decided what, when, and on which evidence.
What happens when new evidence arrives?
The signed decision is flagged for re-review. Nothing stays quietly approved on evidence that has since changed, which is the failure mode a gate is supposed to prevent.
Can we disagree with the recommendation?
Yes, and the design assumes you will. Every input is exposed with its source specifically so a disagreement can be about an assumption rather than about the conclusion.
Does it work for a stop decision as well as a go?
It is symmetric. A no-go with the evidence written down and signed is a better outcome than a drift, and the memo is structured to support that case as clearly as the other one.
Take a decision that leaves a record.
Bring the program to the gate with a cited six-pillar assessment behind it — and leave with a signed decision that re-opens itself when the evidence moves.