Design the Approval Path Before the Pilot, Not After

An AI approval path is three answers: who signs, what evidence they require, and what format the case arrives in. Design it before the pilot starts, not after. Most AI pilots stall because nobody agreed on the evidence format, and the sponsor is left interpreting a vendor report the CFO would not accept (skillpilotadvisory.ai/about, live 15 Oct 2026).

The approval path is one of four factors the readiness assessment scores. A low score on approval means the pilot will run, produce results, and then wait. The decision does not come because nobody built the case in the format the signer needs.

Three approval-path patterns

Pattern	How it works	When it stalls
CFO-format first	The case arrives as a 12-month model with three scenarios, payback period, and worst case stated plainly, before the pilot starts	Stalls only if the CFO changes. The format survives a person change because it is a document, not a relationship.
Sponsor-to-board	The sponsor builds a one-page summary and presents it at the next board meeting	Stalls if the board meeting is quarterly. The pilot runs for three months with no decision, and the sponsor loses momentum.
Vendor-report relay	The vendor produces a report, the sponsor forwards it, and the CFO interprets it	Stalls immediately. The CFO did not ask for the evidence in this format, and the vendor's credibility is not the sponsor's credibility.

How to design the path before the pilot

1. Name the signer. One person, not a committee. If the signer is a board, name the person who presents to the board.
2. Ask what evidence they require. Payback period, scenarios, worst case. Write it down before you spend anything.
3. Build the case in their format. Hours times loaded cost, 12-month model, three scenarios, payback period, worst case stated plainly. Scan does this in one login (skillpilotadvisory.ai/scan, verified 15 Oct 2026).
4. Set a monthly check date. Not a quarterly review. A monthly date on the calendar keeps the decision alive.

The advisory founding sprint ends in a fund, fix, or stop call, which is the approval path executed with a human in the room (skillpilotadvisory.ai/advisory, fetched 15 Oct 2026). For teams that want to run it themselves, the self-serve path covers the same ground.

What this does not cover

This framework does not cover regulated approvals where a compliance officer must sign before a model touches customer data. That is a separate gate. It also does not cover budget approval, which is a finance process that exists before the AI decision. The approval path here is the evidence-to-decision path for one AI initiative.

Sources and context

• Most AI programmes fail the board test: spend rises while cost, attrition, revenue and cycle time barely move (skillpilotadvisory.ai/about, live 15 Oct 2026).
• The readiness assessment scores approval as one of four factors. A low score means the case is not built in the signer's format (skillpilotadvisory.ai/scan, verified 15 Oct 2026).
• The advisory founding sprint (two weeks, US$3,000) ends in a fund, fix, or stop decision on one AI initiative (skillpilotadvisory.ai/advisory, fetched 15 Oct 2026).
• Malaysia's National Guidelines on AI Governance and Ethics (MOSTI, published Nov 2024) name accountability as an expectation. The approval path is where accountability becomes visible.

FAQ

Who should be the signer? The person who owns the budget line. Not the person who proposed the pilot. If the signer is not the budget owner, the approval is advisory, not binding.
What if the signer wants a different format? Build it in their format. That is the whole point. Ask before you build, not after.
How long should the approval cycle be? Under one month. If the decision waits for a quarterly review, the pilot runs for three months with no decision, and the momentum is gone.
Can the approval path work without Scan? Yes. The path is a discipline, not a tool. Scan automates the evidence production, but the design is the signer, the format, and the monthly date.

Design your approval path with an advisory engagement