Every vendor, board member, and headline is telling you to "do something with AI." Almost none tell you whether your business is actually ready to. An AI readiness assessment answers that honestly — before you spend a dollar on licenses, consultants, or a pilot that quietly stalls in month three.
This guide gives you a transparent, seven-dimension assessment you can score yourself in about five minutes — no 45-minute gated quiz, no form before you see your result. We'll define what an AI readiness assessment measures, walk each dimension with a 1–5 rubric, translate your total into a maturity tier, and — the part almost no other framework does — map every gap to the Microsoft tools you likely already own: Azure OpenAI, Microsoft 365 Copilot, Dynamics 365, Microsoft Fabric, and Purview. By the end, you'll know your weakest link and what to fix first.
What an AI Readiness Assessment Actually Measures
"Is my business ready for AI?" is really seven questions, not one. An AI readiness assessment measures whether your business can realistically succeed with AI right now — a diagnostic, not a strategy. Readiness asks "can we succeed?"; strategy asks "what should we do?" You need both, but readiness comes first: a brilliant strategy on a broken foundation still fails.
And most do fail. A widely cited 2024 RAND Corporation study found that more than 80% of AI projects fail — roughly twice the rate of non-AI IT projects. S&P Global reported in 2025 that 42% of companies had abandoned most of their AI initiatives, up from 17% a year earlier. Treat the figures as directional, but the pattern is consistent: most AI initiatives don't fail because the models are bad — they fail because the business wasn't ready. Dirty or siloed data, a vague use case, weak governance, no adoption plan. The models are good enough; the ground they run on usually isn't.
The good news: readiness gaps are fixable, and knowing yours is the whole point of the assessment. Closing them is the day-to-day work of forward deployed engineering for the mid-market.
The AI Maturity Model: Four Tiers
Most frameworks — including ours — sort businesses onto an AI maturity model with four tiers:
- Nascent — Curious but unprepared. Data is siloed, governance is informal, no clear use case.
- Emerging — Foundations forming. A use case or two identified; data and governance still patchy.
- Operational — AI running in production for real use cases, with governance and measurement in place.
- Leading — AI embedded across the business, actively managed, measured, and scaling.
The goal isn't to be "Leading" before you start — it's to know which tier you're in and the move that takes you up a rung.
The AI Readiness Assessment Framework: Score 7 Dimensions 1–5
Here's how to assess AI readiness in practice: score your business 1 to 5 on each of the seven dimensions below, using this scale.
- 1 — Nonexistent: no capability or awareness.
- 2 — Ad hoc: informal, inconsistent, one-off.
- 3 — Developing: partly in place, gaps remain.
- 4 — Established: repeatable and documented.
- 5 — Optimized: strong, governed, and measured.
Be honest — inflating a score only hides the gap you'll trip over later. Add the seven for a total out of 35.
| Dimension | What a 5/5 looks like | Your Microsoft-stack reality check |
|---|---|---|
| 1. Business alignment & use case | A named, ROI-bearing problem with an executive owner and a target metric | Can you name one process — a quote-to-cash step, a support queue, a finance-close task — where AI would save measurable hours or dollars this quarter? |
| 2. Data foundation | Clean, accessible, governed data; key sources consolidated, not siloed | Is your Dynamics 365 / Business Central data clean and current, and is your wider data estate consolidated (e.g., in Microsoft Fabric) rather than scattered across spreadsheets? |
| 3. Technology & infrastructure | A modern cloud platform ready to add AI services securely | Do you have an active Microsoft 365 and/or Azure tenant, access to Azure OpenAI Service, and somewhere to stand up Copilot Studio or an Azure AI Foundry project without re-platforming? |
| 4. Governance, security & compliance | Documented AI governance; access controlled; risk and compliance owned | Are Entra ID permissions tight (no oversharing), is Microsoft Purview handling sensitivity labels and DLP, and can you map to the NIST AI RMF? |
| 5. People, skills & change readiness | Named champions, an adoption plan, and leadership air cover | Is there a person who owns adoption and a plan to train users — not just a license purchase and a hope that people use it? |
| 6. Process & integration | Documented workflows connected across systems | Are the workflows AI would touch actually documented and integrated, or stuck in manual hand-offs between disconnected apps? |
| 7. Value measurement | A baseline today and a way to prove ROI after deployment | Can you baseline the current cost or time of the target process, so you can prove AI actually moved the number? |
Two dimensions carry most of the mid-market risk, and both are where the Microsoft stack matters most.
Data foundation is where AI projects most often die. You don't need a perfect enterprise warehouse — you need clean, accessible data for your target use case. If your Dynamics 365 or Business Central records are inconsistent, or your data is scattered across spreadsheets and disconnected apps, that's your starting point: consolidating into Microsoft Fabric and connecting the systems your data is trapped in.
Governance, security, and compliance is the dimension teams most often overscore. If you plan to run Microsoft 365 Copilot on your own data, the gating question isn't licensing — it's oversharing. Because Copilot honors the permissions each user already has, loose SharePoint or OneDrive access means it can surface sensitive files to the wrong people. Tight Entra ID permissions, Microsoft Purview sensitivity labels and DLP, and a posture you can map to the NIST AI Risk Management Framework (its Govern, Map, Measure, and Manage functions) are what a real "5" looks like.
Add Up Your Score: Your AI Maturity Tier
Your total out of 35 maps to a maturity tier and a next move:
| Your score (of 35) | Maturity tier | What it means and your next move |
|---|---|---|
| 7–14 | Nascent | Foundations first. A pilot now would likely stall. Fix data quality and basic governance before you buy anything. |
| 15–21 | Emerging | Ready for a scoped pilot on your single strongest use case — while you close data and governance gaps in parallel. |
| 22–28 | Operational | Ready to run AI in production for one or two use cases. Standardize governance and start measuring ROI. |
| 29–35 | Leading | Ready to scale across the business. Focus shifts to optimization, model management, and change at scale. |
One caveat overrides the total: any single dimension scoring 1 or 2 is a blocker. A business that scores 5 on technology but 2 on data governance is not ready to turn Copilot loose on its files. Your weakest dimension — not your average — is your real starting point.
Get Your Free AI Readiness Checklist and Scorecard
We've packaged the rubric above into a free, one-page AI readiness checklist and scorecard you can download and keep — the same 1–5 scale, Microsoft-stack reality checks, and tier bands. Score your leadership team together, then re-run it each quarter. It's the rubric our advisors use in a full assessment, no 45-minute quiz required.
From Score to Roadmap: Closing the Gaps on the Microsoft Stack
A score is only useful if it tells you what to do next. Start with your lowest-scoring dimension and match it to a first move.
| Weakest dimension | Typical first move on the Microsoft stack |
|---|---|
| Business alignment | Run a scoped use-case workshop; pick one ROI-bearing process and define its success metric before touching technology. |
| Data foundation | Clean up Dynamics 365 / Business Central data, consolidate your data estate in Microsoft Fabric, and integrate the siloed systems your data lives in. |
| Technology & infrastructure | Confirm your Azure tenant and Azure OpenAI access; stand up a Copilot Studio or Azure AI Foundry environment for a first build. |
| Governance & security | Tighten Entra ID permissions, deploy Microsoft Purview (sensitivity labels and DLP), and remediate Copilot oversharing before go-live. |
| People & change | Name an adoption owner, build a training plan, and run a small champions pilot before a broad rollout. |
| Process & integration | Document the target workflow end to end and connect the systems it spans. |
| Value measurement | Baseline today's cost and cycle time so the pilot has a number to beat. |
This sequence — data and governance before use cases, use cases before scale — mirrors Microsoft's own Cloud Adoption Framework guidance for becoming "AI ready." It's also the work of our Azure AI package, which turns a readiness score into a delivered roadmap.
Is Mid-Market AI Readiness Different? (Yes)
Most AI readiness frameworks are built for the enterprise — they assume a data-science team, a data lake, and a dedicated ML platform. For a 20-to-500-person business, that framing is intimidating and beside the point.
Here's the honest mid-market version: you don't need an ML department or a data lake to start. You need clean data in the systems you already own, one high-value use case, and enough governance to run it safely. A services firm doesn't have to boil the ocean — it needs its Business Central data to be trustworthy, its Copilot permissions tight, and one painful manual process worth automating.
Right-sizing expectations is itself part of readiness. The teams that succeed start narrow, prove the number, and expand — the ones that stall try to launch an enterprise-grade AI program before cleaning up a single data source.
What a Professional AI Readiness Assessment Costs
A DIY self-score like the one above is free and takes five minutes — enough to start an internal conversation. A facilitated AI readiness assessment goes deeper: an advisor reviews your data estate, tenant configuration, governance posture, and candidate use cases, then validates your tier and hands back a prioritized roadmap.
What does that cost? It depends entirely on scope — how many use cases and data sources are in play, how deep the compliance review goes — so treat any figure as illustrative, not a quote. In our experience, a focused mid-market assessment typically runs from a few thousand dollars to the low five figures over two to four weeks — a fraction of the cost of the stalled six-figure AI project it's designed to prevent.
Graduate to a facilitated assessment when there's real budget on the line, when your data or compliance picture is complex, or when you need an outside expert to align stakeholders around one honest answer.
One aside for when you start building: AI pricing changes often, so confirm current Azure OpenAI Service rates on Azure's official pricing pages before you model costs — don't budget from a number you read in a blog, including this one.
Frequently Asked Questions
What is an AI readiness assessment, and what does it measure?
It's a structured evaluation of whether your business can succeed with AI right now, scored across seven dimensions from data foundation to governance to value measurement. It produces a maturity tier and a ranked list of gaps — measuring readiness ("can we succeed?") as distinct from strategy ("what should we do?").
How do I know if my business is ready for AI?
Score the seven dimensions 1–5 and add them up. A total in the Operational band (22–28) signals you can run a production use case. But the total isn't the whole story: if data foundation or governance scores a 1 or 2, that becomes your first fix regardless of everything else.
What are the dimensions, or pillars, of AI readiness?
Seven: business alignment and use case, data foundation, technology and infrastructure, governance and security, people and change readiness, process and integration, and value measurement. Most credible frameworks share these pillars — what's different here is that each maps to the specific Microsoft tools you likely already own.
Do I need clean data or a data warehouse before starting with AI?
You need clean, accessible data for the use case you're targeting — not a perfect enterprise data warehouse first. A scoped pilot can run well on trustworthy data inside Dynamics 365 or a single Microsoft Fabric workspace. Fix the data your first use case touches; don't wait to consolidate everything.
Are we ready to run Microsoft Copilot on our own data?
The gating question usually isn't licensing — it's oversharing. Because Microsoft 365 Copilot respects the permissions users already have, loose SharePoint or OneDrive access lets it surface sensitive files to the wrong people. Tighten Entra ID permissions and apply Purview sensitivity labels first. Your data stays inside your Microsoft 365 trust boundary and isn't used to train the underlying models.
How much does a professional AI readiness assessment cost?
A DIY self-score is free. A facilitated mid-market assessment typically runs from a few thousand dollars to the low five figures over two to four weeks, depending on scope (treat that as illustrative). You get a validated maturity tier and a prioritized gap-closure roadmap.
The Bottom Line
An AI readiness assessment isn't about proving you're behind — it's about finding the one gap that would sink your first AI project and closing it before you spend the budget. You don't need a perfect score to start. You need an honest one, and the discipline to fix your weakest dimension first.
Score yourself on the seven dimensions above. If your data foundation or governance came in at a 1 or 2, that's your starting line — a solvable problem, usually with Microsoft tools already sitting in your tenant.
Want a second set of eyes on your score? As a Microsoft Partner whose team has 20+ years of combined experience on the Microsoft stack, from Azure to Dynamics 365 and the Power Platform, we at Craftware offer a free AI readiness assessment: we'll review your scorecard, confirm your maturity tier, and send back a prioritized gap-closure roadmap — data, governance, and first use case, in the order that fits your business. Book your free AI readiness assessment and get a clear, honest answer on where you stand and what to do next.
