Readiness check
Program Recovery Readiness Check
Twelve questions to answer before approving another dollar of transformation spend. Work through them with your program leadership. Nothing here requires an email address, a form, or a call.
How to use it
Answer from evidence, not from memory.
Each question has a documentary answer or it does not. Where an answer would have to be reconstructed in a meeting before anyone could give it, mark it unclear. That is itself the finding.
Take the questions in order with the people who would have to produce the answers: the program lead, the sponsor, and the line executives whose operations are meant to change. It takes about ninety minutes.
Count the answers that come back unclear, contested, or reconstructed on the spot. The count matters less than where the unclear answers cluster. Six unclear answers spread across all six themes is a different problem from six concentrated in decision rights.
12 questions · six themes · no email required
The check
Twelve questions, six themes.
Outcomes
A program that cannot state the operating change it will produce is measuring delivery instead of value.
Question 1
What operating change will be true when this program is finished, stated so someone outside the program could verify it?
A weak answer sounds like: A list of systems, deliverables, or milestones rather than a description of how work will happen differently.
Question 2
If the program stopped next month, what would the organization have permanently gained?
A weak answer sounds like: Very little. Value was sequenced entirely to the end, which is a recoverable design error only if it is named early.
Decision rights
Most stalled programs are waiting on a small number of decisions that no individual has authority to make.
Question 3
Which decisions has this program been waiting on for more than thirty days, and who owns each one by name?
A weak answer sounds like: The list cannot be produced quickly, or the owners are committees and forums rather than people.
Question 4
When two workstreams disagree on a cross-functional decision, who settles it without going to the steering committee?
A weak answer sounds like: Everything escalates. If the steering committee meets monthly, every disagreement costs a month.
Workflow ownership
Technical delivery lands on an unchanged operation unless a line executive has agreed to change how their organization works.
Question 5
Which line executive has agreed to change how their organization operates, and what are they measured on?
A weak answer sounds like: The program director is the only named owner. The program has no operating-model owner in the line organization.
Question 6
Who is accountable for adoption, as distinct from delivery, and what are they measured on?
A weak answer sounds like: The same person, on the same measures, with training completion standing in for adoption.
Dependencies
Programs rarely fail inside a workstream. They fail at the boundaries, and boundaries have no reporting line.
Question 7
Name every workstream reporting green whose deliverable another workstream is currently blocked on.
A weak answer sounds like: Nobody can produce this from existing reporting, which means the reporting is not showing enterprise risk.
Question 8
Which external dependency (a vendor, a regulator, another program) could stop this work, and who is managing it?
A weak answer sounds like: The risk is documented in a register but no individual is actively working it.
Operating measures
A program measured against an unverified baseline cannot demonstrate whether it worked.
Question 9
What baseline will this program be judged against, and when was it last verified?
A weak answer sounds like: The baseline comes from the original business case and has not been revalidated since approval.
Question 10
Which assumption in the original business case is no longer true, and when did it stop being true?
A weak answer sounds like: "None." After eighteen months in healthcare, that usually means nobody has checked.
AI and automation readiness
Automating a workflow whose authority model was never written down makes the ambiguity faster, not cheaper.
Question 11
For the workflow you intend to automate, can you produce the decision inventory: which decisions are routine, which are exceptions, and which require clinical or expert judgment?
A weak answer sounds like: No such list exists. Without it the automation scope cannot be defined, only estimated.
Question 12
Take one significant decision from ninety days ago and reconstruct it: the inputs, the rule applied, the approver, and the rationale. Can you?
A weak answer sounds like: You cannot. Adding machine-assisted determinations on top of that will make the audit position worse, not better.
Reading the result
What the pattern of unclear answers usually means.
Treat these as a guide for a leadership conversation. The value sits in which questions went unanswered and why.
0–2 unclear
The structure is sound.
Where a program with clear answers is still behind, the constraint is usually capacity or sequencing rather than design. That is a management problem, and an outside review is unlikely to tell you much you do not already know.
3–5 unclear
There is a structural gap worth naming before the next milestone.
A cluster of unclear answers in one theme locates the problem precisely. Work that theme first, internally, and re-run the check. If the same answers stay unclear after a genuine attempt, the gap is not one of effort.
6 or more unclear
The program is running on assumptions.
At this point additional funding or an automation decision is being made without the information required to make it. An independent review before that commitment is usually cheaper than discovering the same thing two quarters later.
If several answers are unclear, the program most likely needs an independent recovery review before the next funding or automation decision. That review is what the Program Recovery Review is for: three weeks, a fixed fee, and a written answer on what is wrong and what the next ninety days should contain.
Related reading
Where these questions come from.
The check condenses the diagnostic sections of two guides. Both are longer and explain the reasoning behind each question.
Program Recovery Conversation
Discuss one program under pressure.
A structured 45-minute working conversation about one program that is behind, over budget, or about to fund automation on top of an unstable workflow.