elevate techConsulting Group

AI Readiness Assessment

What does an AI readiness assessment include?

An AI readiness assessment scores your company on six pillars: Data, Systems, Process, People, Governance and Strategy. It ends with a scored readiness profile, the three things to fix before you spend on AI, and the first two use cases that would actually pay. A real one takes three to four weeks and is independent of any software vendor.

If you run a company past $10M in revenue, you are probably being pitched AI from every direction. Your software vendors are adding it to their products. Your staff are already using it on their own. The question underneath is simpler: is our business in shape to get a return from it? That is what an assessment answers. This article lays out what is in one, what you should walk away with, and how to tell a real one from a sales exercise. If you want the full service description, see our AI Readiness Assessment page.

What are the six pillars of the AI readiness assessment framework?

Each pillar is a place where AI projects tend to fail. A good assessment looks at all six, because a company that is strong in five and weak in one still stalls on the weak one.

1. Data

AI works on your data, so the first question is whether you have data worth working on. We look at where the important data lives, who owns it, how clean it is, and whether two systems disagree about the same customer or the same order. Most of the companies we see have the data they need. It is just scattered across systems and spreadsheets, and no one trusts it enough to build on.

2. Systems

This is your core stack: ERP, CRM, accounting, operations and any industry software. The question is whether those systems can connect to anything new and whether they are being used the way they were designed. Often the right system is already in place and used wrong. An AI tool bolted onto a poorly used system just produces faster confusion.

3. Process

You cannot improve a process you cannot describe. We map the handful of workflows where AI might help, such as quoting, order entry, collections, scheduling or reporting, and find where the real work differs from the written procedure. Most technology problems are people and process, not the software. AI does not change that.

4. People

Do your people have the skills, the time and the willingness to change how they work? We look at who is already using AI, who is afraid of it, and who would own the result. A tool nobody owns is a tool nobody uses.

5. Governance

Governance is the unglamorous pillar and often the weakest. It covers an AI policy, rules about what data can go into which tools, review of AI output, and vendor and contract terms. Microsoft and LinkedIn's 2024 Work Trend Index found that 75% of knowledge workers use AI at work and 78% of those bring their own tools. If you have no policy, your data is already going somewhere you did not choose. We cover the fix in how to write an AI policy for your employees.

6. Strategy

Last, we ask what AI is for in your business. Not "we should do something with AI," but which outcomes matter: faster quoting, fewer errors, lower cost to serve. Strategy also covers who decides, how projects get funded and how you will know whether one worked.

What do you get at the end of an AI readiness assessment?

You should get three things, in plain language, that an owner or a board can read without a translator.

The output should be board-ready and sponsor-ready. If you are PE-backed, your sponsor will ask what your AI plan is. A scored profile and a short list of funded steps is a better answer than a slide of buzzwords.

Want a quick read on where you stand before you talk to anyone? It takes three minutes.

Get your AI Readiness Score

How long does an assessment take?

Three to four weeks for a company in our size range. The first week is document requests and interviews. The middle is reviewing how work moves through your systems and scoring each pillar. The last week is writing, review with your sponsor, and a readout. It is a fixed-scope piece of work with a defined end, not an open-ended engagement.

Be wary of anything much shorter or much longer. A one-hour online quiz can be a useful starting point, which is why we offer a free one, but it cannot replace interviews. A six-month "AI transformation discovery" is a billing plan.

How do you tell a real assessment from a vendor sales exercise?

Vendor-led strategy is a conflict of interest. The vendor's team works for the vendor, and the "assessment" ends with their product as the answer. Here is how to tell the difference.

QuestionA real assessmentA sales exercise
Who is paid by whom?You pay the assessor. They sell no software.The assessment is free or cheap and the vendor profits from what follows.
What does the output say?Fix these things first, then try these use cases.Buy this platform.
Does it include a "not yet"?Yes. Some companies should wait, and the report says so.Almost never.
Who do they talk to?Executives, operators and IT, across all six pillars.Mainly IT and whoever signs.
Is it tied to a product?No. Recommendations name use cases and conditions.Every finding maps to a feature.

We do not sell software, and that is the point. Our job is to find what fits your business, on your side of the table.

What questions should you ask whoever runs the assessment?

  1. Do you or your firm earn money from any vendor I might end up buying from? Ask for a straight answer, not a reassurance.
  2. Which six or so areas will you score, and can I see a sample of the output?
  3. Who will you interview, and how many hours of my team's time does it take?
  4. What happens if the answer is "you are not ready"?
  5. What exactly do I have at the end, and who can I show it to?
  6. Is the scope and fee fixed before you start?
  7. Will the recommendations name specific products, or use cases and conditions?
  8. Who actually does the work: the person I am talking to, or someone I have not met?

Is there an AI readiness assessment template?

You can build a useful template yourself, and the structure is simple. List the six pillars as rows: Data, Systems, Process, People, Governance and Strategy. For each, write three or four plain questions, such as whether two systems disagree about the same customer, or whether a one-page AI policy exists, and score each pillar green, yellow or red. Add a column for the owner and the first fix.

That gets you a self-score. Our six-pillar self-check lays out what ready and not ready look like for each pillar, and the free AI Readiness Score does the scoring in about three minutes. What a template cannot do is interview the people who do the work or trace how data actually moves, which is where most of the findings come from.

Should you do a self-assessment or a formal audit?

Start with a self-assessment, and move to a formal one when the stakes justify it. A self-assessment is free, fast and good for finding your weakest pillar, but it reflects what you already know and how you see your own company. Self-ratings tend to run kinder than an outsider's.

A formal assessment adds what a self-check cannot: interviews across finance, operations, sales and IT, a trace of how data moves between your systems, a review of the tools in use, including the ones staff brought themselves, and an independent score. Choose formal when a board, lender or sponsor will ask for the result, when a significant AI spend is on the table, or when your self-check shows several reds and you need the order of fixes.

Why does this matter before you spend anything?

Because most AI spending does not pay back. MIT's NANDA initiative, in its 2025 report The GenAI Divide, found that about 95% of organizations saw no measurable P&L return from their generative AI pilots. The tools were not the problem in most cases. The surrounding business was: messy data, unclear ownership, no process change. An assessment is a cheap way to find that out before the pilot, not after.

If you want to go a step further and test yourself, read Is my company ready for AI?, which walks the six pillars as a self-check.

Common questions

How long does an AI readiness assessment take?

A focused assessment for a company in the $10M to $100M range usually runs three to four weeks. Most of that time is interviews, reviewing how work actually moves through your systems, and scoring. If someone quotes a single afternoon, you are getting a survey, not an assessment.

Who should be involved on our side?

One executive sponsor, the people who run finance, operations and sales, and whoever manages IT. Expect about an hour each for interviews, plus a few document requests. It should not pull your team off their jobs for weeks.

Do we need to be using AI already to get an assessment?

No. Companies with no formal AI use often get the most from one, because the assessment shows what to fix before any spend. Many are already using AI informally through staff accounts, which the assessment should surface too.

Should the assessment recommend specific AI products?

No. It should recommend use cases and the conditions needed to make them work. Product selection comes after, and it should be made by someone who is not paid by the product's vendor.

Not sure where to start?

Take a free three-minute score. If a conversation makes sense, it will offer one.