AI readiness assessment: where AI fits in your business, and where it doesn't

Who is an AI readiness assessment for?

An AI readiness assessment is for a business that suspects AI could take work off its team but can’t yet say where. The question may have come from a vendor’s pitch, a competitor’s announcement, or staff already pasting work into AI tools without anyone having decided they should. The assessment swaps those guesses for a ranked list, and it fits three kinds of business differently:

• Established local businesses: the owner and a handful of people who know every task, where a few conversations show which repeating job costs the most time.
• Multi-location and regional businesses: the same work done differently at each site, where mapping it shows what to make consistent before anything is automated.
• National businesses: several departments, many systems and rules about data, where the ranking has to work across teams and say who owns each step.

You can start on your own. The free worksheet below scores one task and tells you whether paying for an assessment makes sense yet.

How do consultants assess AI readiness in businesses?

The published methods we read cover similar ground under different labels. Pearl Technology’s assessment looks across people, process, technology and data, and our free AI readiness assessment worksheet splits it into five parts:

• The work: whether the task can be described in one sentence, is done the same way each time and is written down anywhere.
• The data: where the information lives, how long the last twelve months of records would take to produce, how often they’re wrong and which system wins when two disagree.
• The systems: whether they can be reached through an API or an export, how many are involved and who controls access to them.
• The people: whether one named person owns the task, how the team feels about changing it and who would check the output.
• The economics: whether you know what the task costs today, what happens when it goes wrong and how you’d measure a change.

A worksheet records what you believe about each of those. A paid assessment checks the evidence: it talks to the people doing the work, looks at the records and follows the work through the systems. Where the evidence and the worksheet disagree, the ranking follows the evidence.

What does an AI readiness assessment include?

The assessment we sell is a fixed piece of work with four parts:

• Interviews: conversations with the people who do the work and the people who decide, about how it really runs and where it stalls.
• A map of the work: each workflow written out from where it starts to where it ends, with the exceptions and workarounds nobody wrote down.
• A ranked report: what to automate first, what each opportunity would save and what it would cost.
• A readout: a walk through the report with you, with time for questions.

It doesn’t include building anything. The assessment tells you what’s worth doing, and a build, if one follows, is priced on its own. Because the report ranks what it finds, the bottom of the list matters as much as the top: those are the projects not to start yet.

What is an AI opportunity assessment?

Much the same thing, seen from the other side. An AI opportunity assessment starts from the opportunities: where in the business AI could save time or money, and how much. An AI readiness assessment adds what has to be true first: whether the data is clean enough, the systems can be reached and the team is ready to work a new way. Ours covers both: it maps the work with the data and systems it runs through, then ranks the opportunities by what each would save and cost.

AI readiness on its own means how prepared a business is to use AI: its data, its systems, its people’s skills and its appetite for change. An assessment measures it for the specific work you’d point AI at.

Whichever name a firm uses, ask what its report will rank and how. A useful report puts a saving and a cost beside every idea it ranks.

Is an AI readiness assessment the same as an AI audit?

Sometimes. “AI audit” gets used for two different things. One is what this page describes: a review of how the business works, to find where AI fits. The other is a governance or compliance audit, which checks the AI a business already uses against rules, policies and risks. That second kind is not what we do, and we don’t sign off on compliance. The difference matters because the two end in different documents: a readiness assessment ends in a list of work worth doing, while a governance audit ends in findings measured against a standard.

So when you compare AI readiness assessment services, ask which kind is on offer, what the report will contain and who it’s written for. Our AI readiness checklist is a guide to checking readiness on your own.

How long does an AI readiness assessment take?

Published scopes run from one week to twelve or more, depending on how much of the business is covered. As each publisher’s page gave it on November 2, 2026:

• Main & Machine: its readiness audit takes two to four weeks.
• Entrans: one to three weeks for a single department or location, three to six weeks for a cross-functional assessment, and six to twelve weeks or more for an enterprise with several business units.
• Aries Consulting Group: a focused audit for a mid-market business shouldn’t take longer than three to four weeks, and its enterprise-grade assessment runs four to eight weeks.
• Pearl Technology: it sells a fifteen-day assessment with a fixed timeline, asking a few hours of your team’s time across three weeks.

In the Entrans guide, each longer engagement covers more of the business: more interviews and systems for a cross-functional assessment, and several business units for an enterprise. Your proposal sets out the stages for yours.

What happens after the assessment?

What you do with the report is your call. Three routes lead on from it:

• A roadmap: AI strategy and roadmap work turns the ranked list into a phased plan, with the cost of each phase and the measures that show whether each step worked.
• One deployment, planned in full: AI implementation takes the first opportunity and specifies that one deployment in detail.
• The build itself: AI automation puts that step to work inside your current systems and keeps it running.

Nothing in the assessment commits you to any of them. The AI consulting page sets out how the pieces fit together.

When is an assessment not worth paying for?

When the worksheet already gives the answer. It scores one task from 0 to 60 and puts it in one of four bands:

• Foundations first, 0 to 23: the task isn’t ready, and the fix is groundwork: write the procedure down, bring the data into one place and name an owner. An assessment would mostly tell you the same.
• Ready for one, 24 to 38: one task can be built with careful scoping. Pick the one that takes the most hours and does the least damage when it goes wrong, and keep a person approving its output.
• Ready to sequence, 39 to 50: several tasks are ready, and the open question is their order and what depends on what. This is where a paid assessment or a roadmap earns its cost.
• Ready to scale, 51 to 60: the limit is delivery capacity, usually engineering time or a decision nobody wants to own.

If you already know the task, and the worksheet says it’s ready, you may not need an assessment at all: a 15-minute call about that task can go straight to what a build would involve.

What should you have ready before an assessment?

Not much, but four things make it go further:

• An owner: one person who can make decisions about the work being assessed and get people into the room.
• Examples: a handful of real cases of the work, such as the emails, forms or orders as they actually arrive, with personal details removed.
• Access: read access or exports from the systems involved, so the data itself can be checked.
• Today’s numbers: whatever you already track about the work, such as volumes, hours or error rates, even if it’s rough.

None of it has to be tidy. Finding out where it isn’t is part of the work.

What does an assessment lead to in practice?

An assessment is meant to find work like the email intake we built for Alvy Logistics and DSI Transportation. It runs on AlvyOne, the operations system we built and run with Alvy, and it turns the courier orders both companies receive by email into orders in the dispatch system, over its API. Alvy is owned by our co-founder Noah Alvarez, and DSI is an Alvy company. An order with a gap waits for a dispatcher, with a note on what’s missing. It’s the kind of opportunity a ranked report puts near the top: the work repeats, its inputs need reading but its rules can be written down, and an order that doesn’t check out can be held for a person instead of going through wrong. The AlvyOne case study dates each stage from the first build on.

What a paid assessment adds is the step before a build like that: it looks across the work and ranks it, so you choose the first build with the whole list in view. If you want that view of your own business, a 15-minute call is where it starts, and we’ll say whether an assessment, the free worksheet or a conversation about one task suits you.