It ranges from nothing to several days of an adviser's fees. A self-assessment costs only staff time (Microsoft's online AI Readiness Assessment takes about 45 minutes), vendor-run assessments are often free because a sale follows, and independent assessments are priced by the days of work involved. Ask every provider for the day count and a sample report.
Your staff's time is part of the price too, and often a bigger share than owners expect: interviews, gathering information and a final readout can add up to 15-20 hours across a small team, whatever the adviser charges. Fees vary too widely with scope and adviser for a published range to mean much. What you can do is price every option on the same basis: days of adviser time, hours of your people's time, and what you'll hold at the end.
Four routes to an assessment and what each costs you
An AI readiness assessment is a structured look at whether your business is set up to use AI well: which jobs are worth automating, whether your data and systems can support it, what risks need managing, and what to do first. What an AI readiness assessment involves covers the content. The cost of getting one comes down to who does it and how deep they go.
| Route | Fee | Your staff time | What you get | Watch for |
|---|---|---|---|---|
| Self-assessment with a checklist | None | 1-3 hours | A score and a list of gaps | You're marking your own homework |
| Microsoft's online AI Readiness Assessment | None | About 45 minutes | A position across seven areas, with recommendations | Written with larger organisations in mind |
| Vendor or reseller assessment | Often none | 2-4 hours | Your readiness for their product | The answer tends to be their product |
| Independent paid assessment | Adviser days x day rate | 10-20 hours | Report, process maps, prioritised plan | Quality varies widely; ask for a sample |
| Discovery phase at the start of a project | Part of the project fee | 10-30 hours | Readiness findings plus a scoped first project | May assume the project will go ahead |
The two free options are worth doing first whatever you decide next. A checklist such as the 20-minute AI readiness checklist surfaces the obvious gaps. Microsoft's assessment, hosted on Microsoft Learn, is a 45-minute set of multiple-choice questions across seven areas: business strategy, AI governance and security, data foundations, AI strategy and experience, organisation and culture, infrastructure for AI, and model management. For a 15-person business, the last two areas mostly won't apply, and that's a fair test of any assessment: how much of it is about a business your size?
When a paid assessment is worth it, and when it isn't
A paid assessment earns its fee when the decisions it informs are bigger than the fee. That's usually true when several of these apply:
- You have several candidate processes and no agreement on which comes first.
- You handle sensitive client data and need to know what can safely go into which tools.
- You're about to commit to licences at scale. Microsoft 365 Copilot Business at its $21 list price for 20 users is $21 x 20 x 12 = $5,040 a year; an assessment that shows eight of those people would be fine with the Copilot Chat already included saves $2,016 a year.
- Staff are already using AI tools informally and you don't know what data is going where.
It usually isn't worth it when you already know the one job you want automated, when the business is only a handful of people, or when you haven't yet tried the AI already included in your software. In those cases the money is better spent on building the first thing.
An illustrative self-storage business with five staff shows the small end. Its owner worked through a readiness checklist in an evening and found one clear candidate: answering the same website questions about unit sizes and access hours dozens of times a week. There was nothing for a paid assessment to decide. The owner went straight to trialling a website assistant, and spent the money a review would have cost on the trial instead.
Where the days go in a paid assessment
An adviser's quote is time multiplied by a rate, so the useful question is what the time is spent on. Here's one reasonable breakdown for a firm of 10 to 20 people, in hours. These are estimates for sanity-checking a quote's day count, not a price list.
| Task | Light review | Standard | Deep |
|---|---|---|---|
| Kick-off and reading your documents | 2 | 3 | 4 |
| Staff interviews (45-60 minutes each) | 0 | 6 people: 5 | 12 people: 12 |
| Mapping key processes step by step | 0 | 3 processes: 6 | 8 processes: 16 |
| Data and systems check | 1 | 4 | 16 |
| Review of existing tools and licences | 1 | 2 | 4 |
| Risk, data-handling and policy check | 0 | 2 | 6 |
| Writing the report and prioritised plan | 3 | 8 | 16 |
| Readout meeting | 1.5 | 1.5 | 3 |
| Total | 8.5 hours, about 1 day | 31.5 hours, about 4 days | 77 hours, about 10 days |
Two things follow. A one-day assessment can't include interviews or process mapping, so its findings come from a survey or a workshop and your documents; that's fine for a first view but thin for decisions. And a quote of many days for a small firm should come with a reason, such as several sites, messy data or regulated work, that you can see in the scope.
The review of existing subscriptions deserves its hours. Many small firms buy AI tools they partly own already: Microsoft 365 business plans include Copilot Chat at no extra cost, and Google Workspace business plans now include Gemini. A good assessment finds those before recommending anything new.
Your side of the bill: staff hours
Whatever the fee, your people spend time on the assessment. For the standard scope above, at a 14-person business:
- Owner or manager: kick-off, answering questions, readout, about 5 hours
- Six staff interviewed: about 1 hour each including preparation, 6 hours
- Office manager gathering documents, access and sample data: about 4 hours
- Readout attended by three people: 1.5 hours each, 4.5 hours
That's 19.5 hours. At an illustrative blended cost of $35 an hour, about $680 of staff time, before the adviser's fee. Cheap formats aren't always cheap on this line: a 90-minute workshop with eight staff costs 12 staff hours on its own.
A staff survey before the interviews saves time on both sides. An assistant can summarise the answers. The prompt one estate agency used, with names removed first:
Below are anonymised answers from 14 staff to five questions about
repetitive work, tools they use, and where mistakes happen.
Group the answers into themes. For each theme give: how many staff
mentioned it, the roles involved, two short direct quotes, and the
hours per week mentioned if any. Don't recommend solutions.Part of the summary (illustrative):
Theme: re-typing property details between systems
Mentioned by 9 of 14 (sales negotiators, lettings, admin)
Quotes: "same details into three places"; "portal listing done twice"
Hours mentioned: 2-5 per week per person
Theme: chasing landlords for documents
Mentioned by 4 of 14 (lettings)
Quotes: "safety certificates always last minute"The summary was accurate but merged two different complaints under re-typing: listing portals and copying offers into the sales progression spreadsheet. The adviser split them in the interviews, because they have different fixes. That's the right division of labour: the assistant sorts, a person who asks follow-up questions interprets.
Comparing three quotes for a 14-person estate agency
The illustrative business here is an estate agency with sales and lettings teams, 14 staff, Microsoft 365, a property CRM and listings on the main portals. It requested three quotes against the same short brief (how to write a request for proposal covers writing one), and asked each provider for an anonymised sample report from a similar-sized client. The sample reports proved more useful than the quotes. The brief was short:
We are a 14-person estate agency (sales and lettings) using Microsoft 365,
a property CRM and the main listing portals. We'd like an AI readiness
assessment covering: where staff time goes on repetitive work; whether
our data and systems could support automation; which existing subscriptions
we don't use; risks in how staff use AI today; and a prioritised
90-day plan. Please quote the number of adviser days, what our staff
will need to provide, the deliverables, and include an anonymised
sample report from a similar-sized client.Asking every provider the same questions is what made the answers comparable. Three replies came back:
| Quote A | Quote B | Quote C | |
|---|---|---|---|
| Provider | Independent adviser | Independent adviser | Software reseller |
| Adviser time | 1 day: survey plus a 90-minute workshop | 4.5 days: interviews, 3 process maps, data and licence review | 2 hours: a call and a questionnaire |
| Fee basis | Fixed fee for 1 day | Fixed fee for the defined scope | No fee |
| Staff hours needed | 14 (workshop of 8, plus owner) | 19.5 | 4 |
| Deliverable | Slide deck of AI ideas | Report, maps, prioritised list, 90-day plan | Licence proposal |
| Actionable items in the sample report | 1 (the rest generic) | 6, each with an effort and value estimate | 1: upgrade licences |
| Adviser days per actionable item | 1.0 | 0.75 | Not comparable |
At the same day rate, Quote B costs four and a half times Quote A, yet it costs less per item the agency could act on, and its staff-hours line is only a little higher. Quote C cost nothing and was useful for exactly one question, whether the agency needed the paid Copilot licence, which is a question worth asking a reseller as long as you know that's what you're getting.
The agency chose B and negotiated one change: two process maps instead of three, with the third, the lettings renewal process, as an optional extra day. The quote came down in proportion to the hours, which is only possible when a quote is built from hours in the first place. Fixed-scope versus hourly consulting explains why a fixed fee for a defined scope suits this kind of work.
What the fee should buy: seven deliverables to insist on
- A prioritised list of opportunities, each with a rough effort, value and risk, not just a list of ideas.
- Maps of your most important processes, showing where the time goes and where AI would fit.
- Findings on your data: where it lives, how clean it is, and what must be fixed before any automation.
- A review of your existing subscriptions, including AI features already included in your plans.
- Risks and policy gaps: what staff may paste into which tools, and what customer-facing AI would need.
- A 90-day plan with named owners, starting with one project small enough to finish.
- Everything in your hands: the documents, maps and data summaries are yours, with no follow-on contract required to use them.
Add an eighth if you can get it: a short list of things not to do yet. A good assessment kills weak ideas as well as recommending strong ones.
Here's how the absence of these shows up. A wholesaler paid for a one-day review and received a report recommending it "adopt a modern CRM with AI" and "centralise data into a data lake". It already had a CRM with AI features it hadn't switched on, and it had no need for a data lake. The report never named a single system the business used. That's the test to apply to any sample report: does it mention the client's actual tools, people and numbers? If not, you're buying a template.
Free assessments: who pays, and what that means for the advice
A free assessment is paid for by something. Usually it's the sale that follows: a software licence, a partner's implementation project, or an adviser's paid engagement. That doesn't make the advice wrong, but it shapes it. Before accepting one, ask:
- Who funds this, and what are you hoping I'll buy afterwards?
- Will you recommend a competitor's product, or no product, if that's the right answer?
- What will I receive in writing, and can I use it with another provider?
- What won't you look at?
Answered openly, those make a free assessment useful. A reseller that says "we'll check whether Copilot fits your team and tell you if Copilot Chat is enough" is giving you honest value. For free sessions with advisers generally, what you get from a free AI consultation covers how to use them well. These questions apply to any adviser, me included.
Budget for what the assessment recommends
An assessment that recommends five projects has created five budgets, so decide before it starts how you'll judge its findings. The estate agency used its own survey as the yardstick. The biggest theme, re-typing property details between systems, was reported at 2-5 hours a week by nine people. Even at a cautious one hour a week each, that's 9 x 48 = 432 hours a year, about $15,000 at $35 an hour. That made it the obvious first project, and gave the agency a way to tell, a year later, whether the assessment had paid for itself: not by the report's length, but by whether those hours came down.
Keep the first project small enough to finish within the 90-day plan, and leave the rest of the list for later. Findings age quickly when tools change monthly, so an assessment's plan is worth revisiting after the first project rather than executing in full.
Spending less without losing the useful parts
- Do the free steps first. A checklist and Microsoft's 45-minute assessment clear up the basics, so paid days go on the harder questions.
- Narrow the scope. Two or three processes that matter most, not the whole business. The wholesaler above later ran its own checklist, then paid for one targeted day on order processing alone, and got more from it than from the broad review.
- Prepare the material. A list of your tools and licences, a rough process description and a few sample documents save adviser hours; preparing for an AI consultation has the list.
- Run a survey before interviews. Shorter interviews, better questions.
- Fold it into discovery. If you already know roughly what you want to build, a paid discovery phase can combine the readiness check with scoping the first project.
The cheapest assessment is the one that ends in a decision you act on within a month. Price quotes by days, count your staff hours, read a sample report, and pay for the version that names your systems, your people and your numbers.
Further reads
- How to Audit Your Workflows for AI Opportunities Yourself — Do part of the assessment yourself before paying anyone.
- AI Audit vs AI Readiness Assessment: What's the Difference? — Make sure you're buying the right kind of review.
- Signs Your Business Isn't Ready for AI Yet and What to Fix First — Fix the obvious gaps yourself before paying anyone.
- How Much Does an AI Consultant Cost for a Small Business? — What advisers cost for work beyond an assessment.
- AI Consultant Red Flags: 12 Warning Signs to Walk Away From — Warning signs in any adviser's pitch.
- After an AI Consultation: Turn the Advice Into a 30-Day Plan — Turn the findings into action in the first month.
- AI Tools and AI Development: The Complete 2026 Guide — the AI hub, including every tutorial in the AI-for-business series.
Sources: Microsoft Learn (AI Readiness Assessment description); Microsoft 365 business plan details on included Copilot Chat; Google Workspace plan details on included Gemini. Checked September 2026.