Reading your org's AI readiness honestly

Every vendor will tell you your organization is AI-ready. A diagnostic that isn't flattering gives you a clearer picture of where the friction will be.

The five readiness dimensions

Score each 1-5. The low scores tell you where the work is.

1. Data readiness. Is your data in shape to be useful for AI?
2. Tooling readiness. Do you have modern SaaS infrastructure AI can connect to?
3. Permission readiness. Do you know who can see what, and is it enforced?
4. Leadership readiness. Is senior leadership aligned and engaged, not just approving?
5. Workforce readiness. Are teams prepared to adopt new workflows?

Data readiness

Questions:

• Is important data digital (vs. paper or PDFs of PDFs)?
• Is it structured and searchable?
• Are there canonical versions of documents, or are they duplicated everywhere?
• Are data access controls actually enforced (not just documented)?

Score 5 if most content is digital, structured, de-duplicated, with enforced permissions. Score 1 if much of your critical content lives in email attachments and version-numbered Word docs on shared drives.

Most organizations score 2-3 here. Fixing it is multi-year work; AI adoption accelerates when it's 4+.

Tooling readiness

Questions:

• Are your productivity tools cloud-based (M365, Google Workspace)?
• Do major systems have APIs?
• Is SSO standard?
• Is IT equipped to evaluate and adopt new SaaS quickly?

Score 5 if cloud-first, API-first, modern identity, fast IT approvals. Score 1 if on-prem legacy systems dominate and every new tool takes 6+ months to approve.

Permission readiness

This is often the hidden blocker. Questions:

• Do you know who has access to what content?
• Are permissions enforced in source systems?
• When someone leaves, is their access systematically removed?
• Can your identity provider enforce role-based access across systems?

Score 5 if you can produce an access map on request and it's accurate. Score 1 if "I don't know who sees this document" is a common state.

AI tools that respect permissions need permissions to exist and be accurate. Many orgs discover during AI rollout that their permissions are a mess.

Leadership readiness

Not "does the CEO endorse AI?" — they always do. Actual questions:

• Is there a named AI program owner at the exec level?
• Is there a budget that accounts for tools, integration, governance, change management?
• Are executives visibly using AI themselves?
• Is there a cross-functional committee (IT, legal, HR, business units)?

Score 5 if all four are clear yes. Score 1 if AI is "the CEO's pet project" with no one operationalizing it.

Workforce readiness

Questions:

• Has the workforce used modern SaaS productively before?
• Is there a culture of learning new tools?
• Is there a baseline of comfort with AI (ChatGPT-use, general literacy)?
• Are managers able to coach team members through workflow changes?

Score 5 if the workforce has demonstrated ability to adopt new tools effectively. Score 1 if past rollouts have struggled and digital literacy is uneven.

What to do with the score

• All 4-5: You're ready for ambitious rollouts. Move.
• Mixed 3-4: Roll out selectively; pilot in ready areas first; invest in the weak dimensions.
• Mostly 2-3: Start narrow; fix the foundations; expect slow expansion.
• Predominantly 1-2: Don't do a big AI program yet. Fix data, tooling, permissions, or leadership first.

Organizations that ignore a 1-2 score tend to produce expensive pilots that don't scale.

The uncomfortable questions

• Do leaders actually use AI tools themselves, or just endorse them?
• Is our data really as organized as we'd want to tell a vendor?
• Are we conflating "we use Copilot" with "we have AI strategy"?
• If we're honest, why hasn't our previous technology transformation stuck?

These questions are uncomfortable. They're also the right ones.

Fixing dimensions

Data: multi-year effort; start with the top 5 sources that matter for AI use cases. Tooling: 6-18 months; modernize procurement and IT processes. Permissions: 3-6 months of concentrated work; painful but necessary. Leadership: can improve in months with the right sponsor; can also stagnate if misaligned. Workforce: variable; good change management accelerates; takes years to fully shift.

The sequenced approach

Most orgs benefit from:

1. 6 months: data cleanup + permissions audit + leadership alignment.
2. 6 months: narrow AI pilots in high-readiness areas.
3. 12-18 months: expansion based on what the pilots taught.

Compressing this timeline usually backfires.