AI Pre-Flight Checklist

1

What's the business problem, not the tech wish?

Can I clearly name a single pain point or KPI I want to move in the next 6–12 months? Is it worth solving even without AI?

I have a clearly defined business problem with measurable impact

I have a general idea but need to define it more precisely

I'm exploring AI because it's trendy, not for a specific problem

2

Do we have a measurable success target?

What does "done" look like? Can we measure it in hard numbers (time saved, revenue gained, errors reduced)?

Clear metrics and success criteria are defined

Some metrics identified but need refinement

No clear success metrics defined yet

3

Is my data accessible and trustworthy?

Do we know where the relevant data lives and who owns it? Is it clean, current, and legally safe to use?

Data is organized, accessible, clean, and compliance-ready

Data exists but needs cleanup or better organization

Data is scattered, poor quality, or legally unclear

4

Do I have the people and time to run a pilot?

Who will own the project and be accountable for results? Who will review and approve AI outputs during early stages?

Dedicated project owner and review team assigned

Some resources available but need to formalize roles

No clear ownership or available bandwidth

5

Can we afford to learn?

What's my maximum budget for a 6–8 week experiment (time + money)? Am I comfortable if the pilot only produces learning, not ROI?

Budget allocated with expectation of learning over immediate ROI

Some budget available but need to manage expectations

Budget requires immediate ROI or is severely limited

6

Are there low-risk, high-impact use cases to start with?

Can I find at least one process that's repetitive, rules-based, and measurable? Will automating or assisting in this process create visible wins for the team?

Multiple low-risk, high-impact processes identified

Some potential processes but need validation

Struggling to identify suitable starting processes

7

Do we understand the risks?

What's the worst-case scenario if the AI makes a mistake here? Will it damage customers, compliance, or brand trust?

Risks thoroughly assessed with mitigation strategies in place

Basic risk assessment done but need more detailed planning

Haven't fully considered potential risks and consequences