AI Readiness Assessment Services: What They Actually Cover

An AI readiness assessment is a structured look at your operation that answers one practical question: where would AI pay off here, and what has to be true before it can? A good one starts with the work itself, then looks at the data, systems, people, rules, and direction around that work. It finishes by making a call on each workflow it examined: go, wait, or no.

That last part is the one most services skip. AI readiness assessment services range from a five-minute online quiz to a twelve-week consulting engagement, all sold under the same name. Of the 13 top-ranking pages we reviewed in September 2026, 7 promised a score, a scorecard, or a maturity level. The trouble is that a score ranks you against other businesses without naming the workflow you should change first, which is the reason you went looking for help.

The need is real, and it's sharpest for smaller operations. Across OECD countries in 2024, only 11.9% of firms with 10 to 49 employees used AI, against 40% of firms with 250 or more (OECD, AI Adoption by Small and Medium-Sized Enterprises, December 2025). Our reading is that the gap has less to do with access to tools, which anyone can buy, and more to do with knowing where in the business those tools would actually fit.

This page covers what an assessment should examine, how one runs, what you'll need to prepare, and what the result should look like. We should disclose something first. Avolis runs diagnostics for a living, which gives us a stake in how you read this, so we point to a free five-minute self-assessment here and to our consulting-firms page for the cases where you don't need us at all.

Key Takeaways

• An AI readiness assessment should end in a go, wait, or no decision on named workflows, because a maturity score on its own doesn't tell you what to build.
• Small firms trail badly: in 2024, 11.9% of OECD firms with 10 to 49 employees used AI, against 40% of firms with 250 or more (OECD, 2025).
• "Not applicable" was the top reason US firms gave for not planning to use AI in 2023–24, at 80.9%, and Census researchers note that some firms lacking knowledge of AI may have answered that way.
• Only 2.2% of those firms cited a lack of data, so data is rarely the reason firms give for staying out.
• Of small employer firms using AI, just 7% had fully integrated it into the business, and about half were still experimenting (Federal Reserve Banks, 2026).
• Published assessments run from five minutes (OECD's free tool) to twelve weeks, and the real difference between them is whether anyone measures your actual work.

Table of Contents

• What is an AI readiness assessment?
• Why readiness matters more at small firms
• What a readiness assessment should examine
• How AI readiness assessment services run
• What to prepare before it starts
• The result should be a decision, not a score
• Readiness looks different at 10 to 200 people
• What a self-assessment can and can't tell you
• What readiness predicts about the result
• Where Avolis fits
• Frequently Asked Questions
• Continue Learning

What Is an AI Readiness Assessment?

An AI readiness assessment for business is a review that identifies where AI would create value and what has to change before it can. It should answer three questions in order. Where does work drain time or money today, and which of those drains can AI actually change? Then, what has to be true first (in data, systems, people, and rules) for that change to stick? The order is the method. You can't judge what has to change until you know which work you're changing, and you can't pick the work until you've seen where the hours go. Most frameworks we reviewed start with the third question, which is backwards for a small operation, since the answer to it depends entirely on the first two.

People use readiness and maturity interchangeably, but they measure different things. Maturity describes how advanced an organization is with AI overall, usually as a level on a scale. Readiness describes whether a specific move is ready to make right now. In other words, a 30-person HVAC company can sit near the bottom of every maturity scale and still be ready to automate one job. Think of the quote follow-up calls its office manager makes by hand every Friday afternoon. Our guide to the AI maturity model covers the level-based view.

As of September 2026, there's no standard for what AI readiness assessments include. The same label covers very different products, and the biggest difference between them is how much time anyone spends looking at your work:

Durations and deliverables come from each provider's own published pages, and the last row is our own format, so weigh it accordingly. None of these formats is wrong for everyone, and the self-scored tools are useful enough that we'll come back to them later. What's worth noticing is how the question changes as you move down the table, because the early formats ask what you think about your business and the later ones go and check, which is the difference between describing an operation and measuring one.

Why Readiness Matters More at Small Firms

Smaller firms adopt AI at a fraction of the rate of large ones, and the gap is widest exactly where operations-heavy businesses sit. In the OECD's December 2025 paper for the G7, AI use among firms with 10 to 49 employees was 11.9%. Firms with 50 to 249 employees reached 20.4%, and firms with 250 or more reached 40%.

Sector makes the picture sharper. The same paper puts AI use at 7.2% for construction firms, the lowest of the sectors it names, against "almost 45%" in information and communications. That puts a 40-person contractor in two low-adoption groups at once. We don't read that as a verdict on the business. It suggests nobody has yet mapped where the tools would fit.

The reasons firms give for not using AI point the same way. When the Census Bureau asked US businesses that didn't expect to use AI why not, the answer was overwhelmingly about relevance, and cost and risk barely registered.

Of the firms not expecting to use AI, 80.9% said it wasn't applicable to their business, and just 7.3% cited a lack of knowledge about what AI can do. The Census researchers flag the catch themselves, noting that some firms lacking that knowledge "may have responded as AI being not applicable" (Bonney and others, Tracking Firm Use of AI in Real Time, Census CES-WP-24-16R, revised October 2024).

That caveat is the job of a readiness assessment in a nutshell. "Not applicable" is a conclusion, and sometimes it's the right one. Our suspicion is that it's often reached by thinking about the trade work itself (the roofing, the machining, the showings) rather than the intake, estimating, and scheduling around it. An assessment is how you find out which part of the business the conclusion is actually true of.

Our reading of the data: look at the bottom bar. Only 2.2% of firms not planning to use AI cited a lack of data. So the most repeated advice in this category, clean up your data first, addresses a blocker almost no non-adopter reports. What they do report, "not applicable," may partly mean they can't see where AI fits, as the researchers themselves caution. That's a question about your workflows, and you can usually answer it with the data you already have.

What a Readiness Assessment Should Examine

A complete assessment examines six areas, and the first is the one most services leave out, which is the work itself. Published frameworks tend to converge on the other five. AI Singapore's index, for instance, scores leadership and culture, ethics and governance, business value, data foundation, and infrastructure (AI Singapore, AI Readiness Index), and most consulting firms' pillar lists look much the same. Few of them have a pillar for measuring how the work runs today: of the 13 top-ranking pages we reviewed, one service page, Mindcore's, listed process automation potential as its own area.

The order matters more than the list. If you start with data, the way most checklists do, you'll find problems, because every business has them. You can then spend the first month cleaning records that no workflow was ever going to use. If you start with the work, you find out which data problems matter to the workflow you're actually changing, and in our diagnostics, most of them don't.

People deserve more weight than they usually get. In the OECD's survey for Generative AI and the SME Workforce (2025), covering four G7 countries, 50% of SMEs said their employees lack the skills to use generative AI, the kind that drafts text and answers questions in plain language. We cite that figure as the December 2025 G7 paper reports it. Treat it as a scoping input rather than a reason to wait. A workflow your office lead can run after a morning's training is ready now. One that depends on a skill nobody on the team has isn't ready yet, however promising it looks on paper.

The changes firms make around AI are mostly about work and people. In the Census snapshot, the most common adjustments among AI users were training existing staff (20.8%) and developing new workflows (19.7%), while only 8.0% changed how they collect or manage data. Among firms expecting to adopt, 41.9% anticipated training staff and 37.6% anticipated developing new workflows (all firm-weighted), so an assessment that stops at technology misses the changes firms themselves report making.

For the full framework behind each area, and how the areas are weighted, see our AI readiness assessment methodology.

How AI Readiness Assessment Services Run

A useful assessment runs in four steps: scope, measure, score, and decide. The difference between a good one and a thin one is almost entirely in the second step. That's where someone finally stops asking about the work in general and starts measuring it with the people who do it, which no questionnaire can do for you.

To make the steps concrete, picture a 45-person commercial roofing contractor whose owner suspects the office is the bottleneck but can't say exactly where. It's a composite we've put together for illustration, not a client, and we'll follow it through to the result.

1. Scope. The assessment opens with one conversation with the owner and whoever runs operations, and the goal is to pick two or three workflows to examine closely rather than to survey everything. Even a small business runs more core workflows than anyone can examine properly in a couple of weeks, and looking at all of them lightly tells you less than looking at three well. For the roofer, those three might be bid preparation, crew scheduling, and the invoicing that waits on job-completion photos from the crews.

2. Map and measure. Someone works through each workflow with the people who do it, counting how many bids, intakes, or work orders go through in a week and timing the steps. They note every place a job waits on a person, a phone call, or a field that has to be retyped from one system into another. This is where the business case comes from. For the roofer, it's where the estimator might show you that every bid starts with copying roof measurements out of one report into a spreadsheet, and then again into the proposal.

It's also where the owner's estimates get tested. Across our own diagnostics, the owner's guess at how long a back-office task takes is usually low. (Our own client work, not independent research.)

3. Score. Each workflow gets rated on two things: the impact if it changes, and the effort it takes to change it. The high-impact, low-effort work goes first. We score every process in scope this way, and it's the one step we'd insist on from anyone you hire.

4. Decide. The output is a ranked list of workflows, each with a recommendation and the reason for it, which we cover in more detail below.

At Avolis this runs over about two weeks, before we recommend any tool. AI readiness assessment consulting engagements elsewhere often take longer, and some of that time is real work: Brewster, for example, publishes that its assessment "includes up to 40 stakeholder interviews" at its largest scope (Brewster Consulting Group, AI Readiness Assessment). At 10 to 200 people you rarely need that many conversations, and in our diagnostics, the right three people, walking through the work as they actually do it, tell you more than a round of interviews with everyone.

For a closer look at the interview questions, scoring rubrics, and what consultants look for in the work itself, read how consultants assess AI readiness in businesses.

What to Prepare Before It Starts

Most of the preparation is about access rather than documents. None of the 13 top-ranking pages we reviewed tells you what to have ready, so here's the list we'd hand you. In our experience most of it takes an afternoon to pull together, and it saves a lot of back-and-forth later.

• A few months of records from the systems that run jobs. That means exports from your CRM, field service software, property management system, or job-costing tool, covering estimates sent, jobs won, and work orders closed.
• The spreadsheet. Every operation has one that isn't in any system, and it's usually the one that matters most.
• Read-only logins. You want enough access for someone to see how work moves, and not enough for them to change anything.
• Time with the people who do the work. Managers matter, but the person who answers the phone is the one who knows where intake breaks.
• One owner for the process. Someone has to make the call when the assessment asks whether a step is actually needed.
• Your rules. Pull together anything a customer contract, insurer, or regulator says about where data can go.

You don't need clean data, a written AI strategy, or anyone technical on staff, and that's the whole point of doing this first. If you'd rather work through a structured version yourself, our AI readiness assessment checklist turns the same areas into questions.

The Result Should Be a Decision, Not a Score

The single most useful thing an assessment can hand you is a decision on each workflow it examined: go, wait, or no. Scores and levels are fine as context, but a decision is the thing you can actually act on the following week.

For the roofing contractor, a plausible result would be go on bid preparation and wait on crew scheduling until the crew leads agree on one way of logging availability. Invoicing would be a no, because the accounting software the office already pays for can pull the completion photos straight from the crews' field app once the two are connected, which is a setup task rather than an AI project. That's still the illustrative composite, and we've deliberately given it one of each outcome so that all three are visible, but a real assessment's mix depends entirely on what the measurements show, and it's common for a business to come away with several waits and only one go.

A "no" list is evidence that someone measured. If every workflow comes back "go," it's worth asking what was ruled out and why, because an assessment that finds nothing to rule out probably didn't look very hard. When a diagnostic finds nothing worth automating at all, the honest answer is to say so, and in our diagnostics we do.

Even AI Singapore's index agrees that the top of the scale isn't the goal. It runs from level 0, "AI Unaware," to level 4, "AI Catalyst," and it names level 2, "AI Ready," as the recommended target for most organizations. In other words, what counts is being ready for the specific thing you're about to do.

If you're not sure your business should be doing any of this yet, start with should we implement AI, which is the decision that sits above all of these.

Readiness Looks Different at 10 to 200 People

Most AI readiness assessments were built for large companies, and they don't scale down well. Pillars like model management, machine-learning operations, and enterprise governance assume a data team that a 30-person contractor doesn't have and doesn't need, and three adjustments make the difference.

First, governance shrinks to a page. You need a short written policy on what customer and employee information can go into which tools and who approves new ones. One owner can keep it current.

Second, process discipline counts for more than technology. A Census study of US manufacturers found that industrial AI use hurt productivity and profitability in the short run, a dip before any gain. Among older establishments, abandoning structured production-management practices accounted for roughly one-third of those losses (McElheran, Yang, Kroff, and Brynjolfsson, The Rise of Industrial AI in America, Census CES-WP-25-27, April 2025). The lesson for a small operation is a blunt one: keep the checklists, the job reviews, and the routines that already work, and make sure the assessment checks that you'll still have them after the build, because a new tool is far easier to swap out than the habits that kept jobs on schedule before it arrived.

Third, depth beats breadth, because a small operation's readiness shows up in two or three workflows rather than across six organizational pillars.

Small firms also start from a different place. Among small employer firms using AI, about half were still experimenting with it, according to the Federal Reserve's 2025 survey of 6,525 businesses (Federal Reserve Banks, 2026 Report on Employer Firms, March 2026).

Another 44% had partially integrated AI into business processes, and just 7% had fully integrated it into the business. So a small firm that already uses AI usually isn't deciding whether to start; it's deciding which of its experiments deserve to become permanent. Our page on AI readiness assessments for SMBs goes further into that starting point.

What a Self-Assessment Can and Can't Tell You

A self-assessment is the lightest format in the table above, and it's a sensible first step. Two good ones come from public or nonprofit bodies, and neither is trying to sell you a follow-up call.

The OECD SME AI Readiness Tool takes about five minutes and stores no data, since your answers are processed in your browser (OECD, SME AI Readiness Tool). It maps your business to a readiness profile and lists government support programs in your country, and it's labeled a pilot that covers G7 countries only, the US included. The AI Singapore AI Readiness Index takes about 20 minutes across 15 questions, costs nothing, and scores you on the five pillars above. Either is a reasonable first assessment for AI readiness.

Both are worth the time, as long as you know what they measure. A self-assessment tells you how you describe your business. It can't time your intake process or notice that your estimator retypes every quote, which is exactly the kind of finding a measured assessment exists for. Free advising routes such as SBDCs and MEP centers are covered on our consulting firms page.

If a self-assessment is as far as you want to go, our guide on how to assess your organization's AI readiness walks through doing it without outside help.

What Readiness Predicts About the Result

The payoff from AI depends on the task and the person doing it, which is why a readiness assessment has to look at both. In our view, most stalled projects were pointed at the wrong workflow, or at the right one without the changes it needed, and a readiness assessment is the step meant to catch both.

The research backs the task-and-person point. In one of the best-measured studies to date, customer support agents with an AI assistant resolved 14% more issues per hour on average. Novice and lower-skilled agents gained 34%, while the most experienced agents saw minimal impact (Brynjolfsson, Li, and Raymond, "Generative AI at Work," Quarterly Journal of Economics, 2025). It was the same tool in the same company, with very different results depending on who was using it. A general maturity score wouldn't have surfaced that, but a look at who does the work might have, and that's a strong case for measuring workflows instead of rating organizations.

Few firms report cutting staff. A 2026 Census and NBER study found that 66% of AI-using firms use it to augment tasks, while employment reductions were rare, at 2% (Bonney and others, The Microstructure of AI Diffusion, NBER Working Paper 35141, April 2026). The same study found that 57% of adopters use AI in three or fewer business functions, so most firms go narrow, and a good assessment is what tells you which narrow.

For the fuller pattern of what goes wrong and why, see why AI projects fail.

Where Avolis Fits

We run a diagnostic first, always, and it's built the way this page describes. We work through the workflows with your team, measure them, score each one by impact and effort, and hand you a ranked map with a go, wait, or no on each. Every engagement with us starts there, and from there we can build the first workflow and stay embedded while your team runs it. Our own AI readiness assessment service is that two-week diagnostic. We built this page's criteria from how we run it, so weigh them with that in mind.

Whether to buy an assessment at all, from whom, and at what scope is covered in our comparison of AI readiness assessment consulting firms, including the cases where you shouldn't buy one and where we're the wrong fit. For the broader partner decision, start with which AI consulting company to choose.

Frequently Asked Questions

What is an AI readiness assessment?

An AI readiness assessment is a structured review that identifies where AI would pay off in a business and what must change first. It examines the work, data, systems, people, rules, and direction, and a good one ends in a go, wait, or no decision on named workflows rather than just a maturity score.

How long does an AI readiness assessment take?

It depends on the format. Free self-assessments take five to twenty minutes, while published consulting engagements run three to four weeks for firms under 50 employees and up to twelve weeks for enterprise-wide reviews. Avolis's own operational diagnostic, focused on a few workflows, takes about two weeks.

What does an AI readiness assessment cover?

An assessment should cover six areas: the work itself, data, systems, people, rules, and direction. Most frameworks cover the last five, but the one most skip, measuring how the work actually runs today, matters most for a small operation, because without it the assessment can't show where time or money is going.

What's the difference between AI readiness and AI maturity?

Maturity describes how advanced a business is with AI overall, usually as a level on a scale, while readiness describes whether a specific change is ready to make now. A business can be low-maturity and still fully ready to automate one workflow, such as quote follow-up or work-order intake.

Do small businesses need an AI readiness assessment?

Often, yes. Small firms adopt AI far less than large ones, 11.9% against 40% in 2024 (OECD, 2025), and a wrong first project costs them proportionally more. Start with a free self-assessment, and move to a measured assessment when you need your workflows timed and ranked.

Continue Learning

Readiness isn't a grade you earn before you're allowed to start; it's a diagnosis of where to start. Good AI readiness assessment services end the same way, by measuring the work, deciding go, wait, or no, and building the first thing that clears the bar.

Before you act on any assessment's "go" list, ask to see its "no" list.

How assessments work:

• How consultants assess AI readiness in businesses
• How consultants assess AI readiness in organizations
• AI readiness assessment methodology
• AI maturity model

Assess it yourself:

• How can I assess my organization's AI readiness?
• How businesses assess readiness for AI adoption
• AI readiness assessment checklist and PDF
• Should we implement AI?

Specific kinds of readiness:

• AI readiness assessment for SMBs
• Data readiness assessment for AI initiatives
• AI infrastructure readiness assessment
• Generative AI readiness assessment
• Organizational readiness framework for enterprise AI

Choosing who does it:

• AI readiness assessment consulting firms
• Which AI consulting company should I choose?
• Why do AI projects fail?

---

Sources

All sources retrieved 2026-09-24.

• OECD, AI Adoption by Small and Medium-Sized Enterprises, OECD discussion paper for the G7, December 2025 (AI use by firm size and sector from the OECD ICT Access and Usage by Businesses database, 2024 or latest year; firms with 10+ employees; unweighted OECD average), retrieved 2026-09-24 — https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/12/ai-adoption-by-small-and-medium-sized-enterprises_9c48eae6/426399c1-en.pdf
• Kathryn Bonney and others, Tracking Firm Use of AI in Real Time: A Snapshot from the Business Trends and Outlook Survey, U.S. Census Bureau Center for Economic Studies, CES-WP-24-16R, revised October 2024 (original March 2024; BTOS AI supplement, September 2023 to February 2024; Table 7 reasons for not planning to use AI; Figure 7a organizational adjustments, firm-weighted), retrieved 2026-09-24 — https://www2.census.gov/library/working-papers/2024/adrm/ces/CES-WP-24-16R.pdf
• Federal Reserve Banks, 2026 Report on Employer Firms: Findings from the 2025 Small Business Credit Survey, March 2026 (fielded 3 September to 14 November 2025; 6,525 firms with 1–499 employees; convenience sample, weighted), retrieved 2026-09-24 — https://www.fedsmallbusiness.org/reports/survey/2026/2026-report-on-employer-firms
• Kristina McElheran, Mu-Jeung Yang, Zachary Kroff, and Erik Brynjolfsson, The Rise of Industrial AI in America: Microfoundations of the Productivity J-curve(s), U.S. Census Bureau Center for Economic Studies, CES-WP-25-27, April 2025, retrieved 2026-09-24 — https://www.census.gov/library/working-papers/2025/adrm/CES-WP-25-27.html
• Erik Brynjolfsson, Danielle Li, and Lindsey Raymond, "Generative AI at Work," Quarterly Journal of Economics 140(2), 889–942, 2025 (NBER Working Paper 31161; 5,179 customer support agents), retrieved 2026-09-24 — https://www.nber.org/papers/w31161
• Kathryn Bonney and others, The Microstructure of AI Diffusion: Evidence from Firms, Business Functions, and Worker Tasks, NBER Working Paper 35141 / Census CES-WP-26-25, April 2026 (2026 BTOS AI supplement), retrieved 2026-09-24 — https://www.nber.org/papers/w35141
• OECD, SME AI Readiness Tool (pilot version; developed under the 2026 French G7 Presidency), retrieved 2026-09-24 — https://sme.oecd.ai/
• AI Singapore, AI Readiness Index (AIRI), version 3.1, retrieved 2026-09-24 — https://aisingapore.org/innovation/airi/
• Brewster Consulting Group, AI Readiness Assessment (published duration, scope, and deliverables), retrieved 2026-09-24 — https://www.brewsterconsulting.io/ai-readiness-assessment
• Mindcore Technologies, AI Readiness Assessment (published duration and deliverables), retrieved 2026-09-24 — https://mind-core.com/services/ai-agents/ai-readiness-assessment/
• Mission Cloud Services, AI Readiness Assessment (published duration and deliverable), retrieved 2026-09-24 — https://www.missioncloud.com/ai-readiness-assessment

On the adoption figures. Surveys measure AI use differently, so their rates don't agree. The OECD and Census figures count AI used in producing goods or services by firms with 10 or more employees. The Federal Reserve counts any use by the business or its employees across firms of 1–499 employees. We've kept each figure with its own source and haven't compared rates across surveys.

On the Census reasons. The 80.9% and 2.2% figures come from a snapshot fielded between September 2023 and February 2024, before the most recent wave of small-business adoption. We cite them because no later Census release publishes reasons for non-adoption at this detail, and because the researchers' own caveat about "not applicable" is the point we draw from them.

On the format table. Durations and deliverables are each provider's own published figures as of the retrieval date. Naming a provider isn't an endorsement, and we haven't worked with any of them. Avolis's two-week figure describes our own diagnostic.

On the page review. "The 13 top-ranking pages we reviewed" means the pages ranking for the primary and secondary search terms of this article, reviewed on 2026-09-24: seven consulting service pages, three vendor assessment tools or quizzes, and three guides. The counts in the introduction, the areas section, and the preparation section come from that set. It's a snapshot of one week's search results, not a market survey.

On the OECD skills figure. The 50% figure comes from the OECD's 2025 Generative AI and the SME Workforce survey, which we couldn't retrieve directly; we cite it as reported in the December 2025 G7 paper listed above.

On first-party claims. Statements such as "across our own diagnostics" describe Avolis's own engagements and internal observation. They are not independent research and are not offered as benchmarks.

---

About Avolis Research Group

Avolis Research Group is Avolis's in-house research practice, focused on how operations-heavy small and mid-sized businesses actually adopt AI. It synthesizes primary economic research, government survey data, and results from real implementations into practical, vendor-neutral guidance.

More about Avolis and how we work · Get in touch