A working diagnostic, free to use

The AI Decision Diagnostic

Answer twelve questions to identify where to focus your AI work. Get scores across four areas and three suggested next steps. Free to use. No email required.

Choose the answers that best describe your organization today. Your results are based on your answers.

DirectionWhere do we start?

If each member of your executive team wrote one sentence on what AI is for in your enterprise, how many different answers would you get?

One, and we make decisions against it.Two or three that overlap.A different answer per function.The question itself would start the debate.

Every downstream decision inherits this sentence. Fragmented intent becomes fragmented spend.

Who owns AI direction today?

A named executive with budget and the authority to say no.A named executive, but without the budget or the mandate to stop anything.A committee that advises, or IT by default.No one. Initiatives start wherever the energy is.

Direction without an owner is a suggestion.

Which best describes your current plan?

A sequenced roadmap: stated first moves, stated deferrals.A list of initiatives without an order.A pilot portfolio that grew on its own.No plan yet, or a strategy document waiting for a decision.

A plan without a sequence makes every initiative first and none of them funded properly.

ValueWhich investments actually deliver a return?

How does an AI use case get selected in your organization?

Scored against value and feasibility, with a named owner per case.Sponsored by whoever asks with the most conviction.Adopted from what peers and vendors announce.We have not formally selected any.

Selection discipline is the difference between a portfolio and a pile.

When a use case ships, do you know where the return shows up?

Yes: tied to a specific financial line and tracked after go-live.Estimated at approval, rarely revisited.Measured in activity, hours saved, not financial outcome.We do not measure it.

Returns that are not accounted for get cut in the next budget cycle, along with the credibility of the program.

How is the work funded?

Deliberately split: run-rate improvements funded in operations, bigger bets funded as investment cases.One central pot for everything labeled AI.Inside departmental budgets, wherever it lands.With whatever is left over.

Optimize and transform are different ambitions. Funded from one pot, the quick wins crowd out the big bets.

ArchitectureDo we build, buy, or blend?

A vendor pitches an impressive AI capability. What happens next?

It is evaluated against criteria we wrote before the pitch, including exit terms and data rights.We pilot it if the demo impresses the right people.We move if a competitor is already using it.It joins a queue that rarely advances.

The buyer with no criteria inherits the vendor's.

The data your first AI use cases depend on is...

Identified, sufficient for the purpose, and owned by someone.Mostly available, but trapped across systems.Of unknown quality until we try.Honestly, we would be guessing.

AI on unexamined data automates the errors you have not found yet.

Build, buy, or blend decisions are driven by...

A stated ownership posture: we know what we must own and what we deliberately rent.Whatever is fastest right now.What our incumbent vendors propose.Case by case, with no consistent pattern.

Ownership decisions compound. Rent your differentiators and you fund a competitor's roadmap.

AdoptionWill our people adopt what we choose?

After go-live, who is accountable for whether people actually use it?

Business leaders, with adoption in their goals.The project team, until handoff.IT and the training calendar.No one, once it ships.

Adoption is a leadership discipline, not an IT rollout. Unowned adoption is a license renewal without a result.

When a new tool lands, how does the way people actually work change?

Workflows are redesigned with the people who run them, before go-live.Training is delivered; workflows are left to adapt.We announce it and expect the tools to be intuitive.The old process and the new tool run side by side, indefinitely.

Tools do not change work. Redesigned workflows do. Capability gaps surface later as quiet non-use.

Do you know how much change your organization is already absorbing?

Yes, and we manage change load as a real constraint, formally or not.Roughly. It is a lot, and nobody measures it.We are saturated. New initiatives compete for attention.We do not track it.

Change capacity is finite and usually spoken for. Sequencing against it is the difference between adoption and fatigue.

0 of 12 answered.

When you want a deeper review

A paid review combines your answers with interviews and records to examine how decisions are made across your organization.

Discovery

The same four areas, examined using a structured intake and three to five executive interviews of 45 minutes each, two weeks elapsed. You receive a decision brief built for the board room and a one-hour discussion: where your team actually disagrees, and the sequence that follows from it.

Scoped once with you, then a fixed fee in writing before any work begins.

The full diagnostic

Discovery, plus your strategy draft, roadmap, and vendor stack under review. Bring the draft; leave with a clear answer, a sequenced twelve-month path, and the funding lanes to carry it. Three to four weeks.

Scoped once with you, then a fixed fee in writing before any work begins.

Connect with M. Pact

You will hear back within two business days.