AI Readiness Assessment: 5-Dimension, 28-Q Framework (2026)

AI Readiness Assessment: A 5-Dimension, 28-Question Framework for Enterprise CTOs (2026)

The most common reason enterprise AI strategies fail is not strategy quality. It's that the strategy was written on top of an organization that wasn't ready to execute it.

Six months in, the data is messier than the strategy assumed. The team can't ship at the cadence the timeline required. The infrastructure can't support the inference load. The change management was an afterthought.

A readiness assessment, done before the strategy is locked, catches these gaps when they cost $20K to address — not $2M after the strategy has been approved and the budget committed.

This article is the 5-dimension, 28-question readiness framework we use with clients before any AI strategy engagement begins. It produces a score per dimension, identifies the 3-5 weakest factors, and tells you what to address before writing the strategy.

The framework comes from running readiness assessments across mid-market and enterprise clients through Internative's AI Studio service.

What an AI Readiness Assessment Actually Covers

A useful assessment evaluates 5 dimensions in 2026:

Data readiness — is your data AI-usable today?
Process readiness — are the workflows AI is meant to improve actually documented?
People readiness — does the org have the skills + capacity to ship and adopt AI?
Culture readiness — does the company actually want to change?
Infrastructure readiness — can the technical stack support what AI requires?

Each dimension scores 0-12 (4 sub-questions × 0-3 each). Maximum total: 60. Anything under 40 is a serious gap that should be addressed before AI strategy execution begins.

Dimension 1: Data Readiness

Q1: Is your data centralized or scattered across systems?

Score 3: Single data platform (warehouse + lake), well-governed. Score 2: Multiple platforms, integration in flight. Score 1: Scattered across 5+ systems, manual joins required. Score 0: No central data infrastructure at all.

Q2: Is your customer/user data clean and well-labeled?

Score 3: Active data governance, regular quality audits, mature labeling. Score 2: Some quality processes, mixed labeling quality. Score 1: Ad-hoc cleaning, no formal governance. Score 0: Data quality unknown / known to be poor.

Q3: Are knowledge bases (docs, KB articles, contracts, policies) digitized and structured?

Score 3: All major knowledge in machine-readable formats with metadata. Score 2: Most knowledge digital, some structured. Score 1: Mix of PDFs, scanned docs, tribal knowledge. Score 0: Knowledge mostly in people's heads or scanned image files.

Q4: Are data access controls in place for AI use cases?

Score 3: Role-based access, audit trails, sub-processor agreements ready. Score 2: Some controls, gaps for AI-specific access. Score 1: Coarse-grained controls, will need rework for AI. Score 0: No formal data access controls.

Data Readiness Score: __ / 12

Dimension 2: Process Readiness

Q5: Are the processes AI is meant to improve documented?

Score 3: SOPs exist, current state mapped, KPIs measured. Score 2: Some documentation, ad-hoc. Score 1: Tribal knowledge, not written. Score 0: Process not formally defined.

Q6: Do you have baseline metrics for processes targeted for AI?

Score 3: Time, cost, error rate, throughput measured monthly. Score 2: Some metrics tracked. Score 1: Anecdotal performance data. Score 0: No process metrics.

Q7: Have you identified specific decision points where AI would add value?

Score 3: 5+ specific decision points identified with current-state cost. Score 2: 2-3 identified. Score 1: Vague ideas, no specifics. Score 0: "We should do AI somewhere."

Q8: Are business owners assigned to processes AI will touch?

Score 3: Named VPs/Directors own each candidate process. Score 2: Some owners, others unclear. Score 1: Ownership disputed or shared. Score 0: No clear ownership.

Process Readiness Score: __ / 12

Dimension 3: People Readiness

Q9: Do you have engineering capacity to ship AI features?

Score 3: ML/AI engineers on staff or budget approved to hire. Score 2: Strong software engineers, no ML specialists. Score 1: Limited engineering capacity overall. Score 0: No engineering team or fully outsourced.

Q10: Is there executive sponsorship for AI?

Score 3: C-level sponsor actively engaged, board awareness. Score 2: VP-level sponsor. Score 1: Mid-level interest only. Score 0: No exec sponsorship.

Q11: Are the people affected by AI trained or planned to be?

Score 3: Change management program designed, training timed to launch. Score 2: Training planned but not detailed. Score 1: Vague plans. Score 0: No training plan — assumption people will "figure it out."

Q12: Do you have someone accountable for AI-specific governance?

Score 3: AI Risk Officer or Council established. Score 2: Shared governance committee. Score 1: Compliance team aware but not active. Score 0: No AI governance defined.

People Readiness Score: __ / 12

Dimension 4: Culture Readiness

Q13: Has the organization successfully adopted new technology in the last 24 months?

Score 3: 3+ major adoptions, lessons documented. Score 2: 1-2 adoptions, mixed results. Score 1: Tried but struggled. Score 0: Strong resistance to new tools.

Q14: How does the org handle automation that affects jobs?

Score 3: Explicit reskilling and role evolution program. Score 2: Case-by-case approach. Score 1: Some friction, undefined approach. Score 0: High fear of automation, no plan.

Q15: Is there tolerance for experimentation and failure?

Score 3: MVPs and failed experiments are celebrated as learning. Score 2: Some tolerance for failure. Score 1: Failure is blame-assigning. Score 0: Risk-averse culture, perfectionist deployment expectations.

Q16: How quickly does the organization make decisions on technology investments?

Score 3: Weeks for $100K decisions, 1-2 months for $1M+. Score 2: 1-2 months for typical investments. Score 1: 3-6 months. Score 0: 6+ months or rarely makes decisions.

Culture Readiness Score: __ / 12

Dimension 5: Infrastructure Readiness

Q17: Is your cloud infrastructure able to support AI workloads?

Score 3: Modern cloud platform (AWS/Azure/GCP) with GPU/inference capacity. Score 2: Cloud-based but limited AI-specific capacity. Score 1: Some cloud, mostly on-prem. Score 0: On-prem-only or legacy infrastructure.

Q18: Do you have observability for production systems?

Score 3: Mature observability stack (logs, metrics, traces, alerting). Score 2: Some observability. Score 1: Limited monitoring. Score 0: No production observability.

Q19: Is your security posture compatible with AI vendor requirements?

Score 3: SOC 2 + ISO 27001 + GDPR/HIPAA as applicable. Score 2: One major framework. Score 1: Working toward compliance. Score 0: No formal compliance posture.

Q20: Can you handle the cost variability of LLM-based systems?

Score 3: FinOps practice in place, cost monitoring + budget alerts. Score 2: Basic cost tracking. Score 1: Cloud costs are a surprise each month. Score 0: No cloud FinOps capability.

Infrastructure Readiness Score: __ / 12

Additional 8 Cross-Cutting Questions

Q21: Do you have a stated AI strategy or are you operating ad-hoc?

Q22: How many AI experiments are running right now across the company?

Q23: What's your sub-processor approval process for new AI vendors?

Q24: Have you defined acceptable use policies for employee AI tool use?

Q25: Is there a budget line item for AI in this year's plan?

Q26: Have you done a vendor lock-in analysis for major AI providers?

Q27: What's your data residency requirement for AI workloads?

Q28: Do you have a board-level AI risk discussion in the last 6 months?

These don't fit the dimension scoring but matter strategically. They surface in the assessment debrief.

How to Interpret Your Score

Total Score 50-60: Ready

You're ready to write AI strategy and begin execution. Focus on use case selection and architecture decisions.

Total Score 40-49: Mostly Ready, 1-2 Gaps

You have most prerequisites in place. Identify the 2 weakest dimensions and address them in parallel with strategy work. Don't let perfect block good.

Total Score 30-39: Significant Gaps

You have gaps in at least 2-3 dimensions. Strategy execution will hit these gaps in months 2-4 and slow down. Better to address them first (Months 1-2) before strategy execution (Months 3+).

Total Score 20-29: Major Foundation Work Needed

Stop. Don't write AI strategy yet. The foundations are not strong enough. Spend 3-6 months addressing data, infrastructure, or culture gaps before AI strategy work begins.

Total Score Under 20: Not Ready

AI work will fail at this readiness level regardless of strategy quality. Begin a 6-12 month readiness improvement program first.

What to Do With the Score

If you scored well in 4 dimensions but poorly in 1

Address the weak dimension in parallel with AI strategy work. Often this is data readiness or change management. These can be improved in 90 days with focused investment.

If you scored evenly mediocre across all 5 dimensions

You don't need AI strategy — you need digital foundations strategy first. Many companies in this state misdiagnose the gap as "we need AI" when the actual gap is "we need to modernize our data and process layer."

If one dimension scores very low (under 4)

This is a blocker. The strategy you write will fail at this dimension. Address it specifically before strategy.

Common Failure Patterns by Weakest Dimension

Data weakest: AI features ship but produce wrong answers because the underlying data is poor quality.
Process weakest: AI doesn't improve outcomes because the baseline process metrics weren't measured, so improvement can't be proven.
People weakest: AI ships but isn't used because no one trained the people affected.
Culture weakest: AI initiatives stall in committee review forever; nothing reaches production.
Infrastructure weakest: AI features ship but cost 5x expected, performance is unpredictable, security audits fail.

What "Internative AI Readiness Assessment" Looks Like

We offer the 28-question readiness assessment as a 3-4 week engagement:

Cost: €12-25K depending on org size and depth
Output: Score per dimension, weakest factor analysis, recommended remediation plan, board summary
Format: Structured interviews + document review + technical infrastructure audit + scoring debrief

Most clients run this before signing for AI strategy work. The readiness gaps surface assumptions that would otherwise show up at month 6 of execution.

The Three Common Mistakes

Mistake 1: Skipping readiness in the rush to "do AI." The pressure to ship AI is real in 2026. But shipping AI on top of an unprepared organization produces a failed deployment that sets the company back 18 months. Spend the 3 weeks on readiness.

Mistake 2: Treating low scores as judgment. Low readiness scores are not failures — they're maps. The organization that knows its data is messy can budget for cleaning. The organization that doesn't know is the one that fails publicly.

Mistake 3: Doing the assessment but not acting on it. The most common pattern: assessment done, gaps identified, then ignored because addressing them is slower than starting AI work. Six months later the gaps cause exactly the failures the assessment predicted. The assessment is only valuable if its findings drive action.

Five Questions That Resolve the Next Step

Have you measured baseline KPIs for processes AI will touch? No — start with process readiness improvement.

Is your data centralized and well-governed? No — data readiness improvement is your highest priority.

Do you have named business owners for each candidate AI use case? No — people/culture work first.

Does your infrastructure support AI workloads cost-effectively? No — infrastructure modernization in parallel with strategy.

Has the board engaged with AI risk and opportunity? No — get executive sponsorship in place before consultant engagements.

Next Step

If you want to run this assessment for your organization, we deliver it as a 3-4 week structured engagement that produces a scored readiness profile and a remediation plan.

Contact: team@internative.net or via internative.net.