Deadline Reality Check
Estimate delivery probability and expose the path to a credible deadline
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 7
- Confidence
- 99%
Deadline Reality Check evaluates whether a project will meet a target date by combining project documents with recent communications. Instead of returning a binary yes or no, the assistant estimates a probability, identifies positive indicators, catalogs delivery risks, and maps the critical path. It then recommends actions that could improve the probability and describes the most likely delivery outcome, including whether only a partial launch is realistic. The final output should expose a concrete leadership choice, such as shipping a limited version on time or delaying for full scope. This makes deadline discussion more honest and actionable. The estimate is still a reasoned forecast rather than a statistical guarantee, so teams must examine assumptions, dependencies, stale updates, and missing evidence before acting.
Origin
Adapted from a Satya Nadella prompt and presented on Marketing Against The Grain as a deadline assessment method for AI-enabled leaders.
Core principles
- 01Express schedule confidence as probability rather than certainty
- 02Balance positive indicators against explicit risks
- 03Identify the critical path before recommending action
- 04Distinguish partial delivery from full completion
- 05Turn schedule analysis into a leadership trade-off
How to run it
- 1
Set the target
State the exact project, deadline, and definition of completion.
Pro tip Separate minimum viable delivery from full feature parity.
Watch out An ambiguous completion standard makes the probability meaningless.
- 2
Gather current evidence
Search project documents and recent communications for progress, dependencies, changes, and unresolved work.
Pro tip Weight recent verified updates more heavily than old plans.
Watch out Stale status reports can inflate confidence.
- 3
Estimate the probability
Ask for an explicit probability of meeting the deadline and the assumptions supporting that estimate.
Pro tip Request a confidence range when the available evidence is weak.
Watch out Do not mistake a generated number for a calibrated statistical forecast.
- 4
Balance signals and risks
List the positive indicators alongside organizational challenges, scope concerns, and delivery dependencies.
Pro tip Flag risks whose owners or resolution dates are unknown.
- 5
Map the critical path
Identify what must be completed, by whom, and by when for the deadline to remain viable.
Pro tip Focus attention on dependencies that can delay multiple downstream tasks.
- 6
Improve the odds
Generate concrete recommendations that raise the probability, such as narrowing scope, resolving a dependency, or reallocating resources.
Pro tip Estimate which intervention has the greatest schedule impact.
- 7
Frame the decision
State the most likely outcome and present leadership with the real trade-off between date, scope, quality, and resources.
Pro tip Name the decision owner and latest responsible decision date.
Watch out Avoid disguising a partial launch as full completion.
In the wild
An assistant reviews project records and determines that progress supports an October release, but scope and organizational risks make full feature parity unlikely. It forecasts a limited functional launch in October and full delivery in November, then asks leadership to choose between those outcomes.
→ Leadership receives an explicit scope-versus-date decision instead of an unsupported promise.
Common mistakes
Asking only whether the date will be met
A binary answer hides uncertainty; request a probability, evidence, critical path, and most likely outcome.
Leaving completion undefined
A team may claim an on-time partial launch while stakeholders expect full functionality, so define the delivery threshold first.
Ignoring probability-improving actions
Risk assessment has little value unless it produces interventions that can change the outcome.
Is it for you?
Best for
Leaders deciding whether to preserve scope, reduce scope, add resources, or move a launch date.
Not ideal for
Projects whose remaining work, dependencies, and progress are undocumented or fundamentally unknowable.
From the transcript
“You want to say what's the probability we'll hit a certain deadline, and then you want to give it the actual project.”
“Recommendations to improve that probability.”
“Key question for leadership, would you prefer limited but functional October launch or a delay to November?”
From the episode
I Stole Microsoft CEO’s 4 Most Powerful AI Prompts