An AI readiness assessment that ends in a ranked list

An AI readiness assessment checks where a company can put AI to work profitably, and whether its data, systems and people can support it. Ours looks at 5 things: the work itself, the data behind it, the systems that hold that data, the people who would use the result and the risk if an output is wrong.

Some assessments end in a maturity score, a number from 1 to 5 across a set of pillars. That is easy to produce and hard to act on. Ours ends in a ranked list of specific opportunities, each priced, with the data checks done. "Your data maturity is 2.4" tells you little. A line such as "invoice entry takes finance 60 hours a month, the PDFs are machine-readable and a pipeline would cost about $14,000 to build" tells you what to do next. (The figures in that line are illustrative.)

We also record which AI tools staff already use, what they cost and what data goes into them. That shapes both the shortlist and any usage policy that follows.

If you want a first pass on your own, our AI readiness assessment checklist lists the questions we ask in week 1. It will tell you whether an audit is worth paying for.

Both versions include interviews, a systems inventory, a ranked opportunity list and a readout. The full audit adds sampled data checks, feasibility tests on real records and a fixed-price proposal for the first build.

The snapshot suits a company that wants to know whether there is enough opportunity to justify more work. Its estimates carry wider ranges, because nothing has been tested.

The full audit suits a company ready to fund a build in the next 3 to 6 months. Feasibility tests remove the guesswork on the top candidates: we take real invoices, tickets or inbound emails and check whether a model or a plain rule-based workflow gets them right often enough for the job. A candidate that fails is marked down, with the reason and what would fix it.

Your data is ready enough when the inputs for a specific task exist in a system we can read, in a form a person could work from. It does not need to be clean everywhere, only for the first use case. What counts as ready depends on the task:

• Workflow automation, such as routing leads or filing invoices, needs structured fields and an API or export. Gaps are fine if a person can fill them today.
• Document processing needs the documents, ideally as digital PDFs. Scans work with more checking.
• A knowledge assistant built with retrieval-augmented generation needs your documents somewhere we can index, with current versions marked. Watch for duplicate policies that give conflicting answers.
• AI agents that take actions need API access scoped to what the agent should touch.

You can run 3 quick tests before talking to anyone. Pick the task, then ask: could we export last month's inputs to a spreadsheet in under an hour? Could a new hire do the task from that export alone? Can we describe what a correct output looks like? If the first 2 answers are no, start with data work. If the third is no, define the task before automating it.

The data readiness scorecard runs these checks properly, on samples from your systems, and prices any cleanup next to the build.

A snapshot audit costs $3,000 to $5,000 and a full audit $7,500 to $15,000, both fixed price.

What moves the price within each band:

• Teams in scope. Each extra team adds interviews, a process map and its own candidates.
• Systems. Pulling samples from a CRM with a clean API is quick. Read access to an on-premise ERP or a legacy database takes longer.
• Regulated data. If samples include health, financial or personal data, we work inside your environment under your controls, which adds setup.
• Feasibility tests. The full audit tests 3 candidates. Testing more is possible and priced per candidate.

Model usage during feasibility tests is billed at cost with no markup; at a few hundred records it is a small line.

For comparison, the Layer3 Labs AI consulting rates guide (October 2026) (opens in a new tab) puts AI discovery and audit work at $2,000 to $5,000, and the Leanware AI consultant cost guide (October 2026) (opens in a new tab) puts small AI strategy assessments at $5,000 to $25,000. Our snapshot sits in the upper half of the first range because it includes a ranked list and a budget estimate for the first build.

Skip the audit if you already know the task and it touches 1 or 2 systems. A team that wants inbound leads qualified and routed into HubSpot needs a build quote, and the Map stage of that build covers the same ground for a single workflow. Ask for a scoping call about sales and CRM automation instead.

The audit is also the wrong fit when:

• You need a company-wide AI strategy with governance, board reporting and a usage policy. That is corporate AI consulting, which adds pilots and a 12-month roadmap.
• Your core systems are mid-migration. If the CRM or ERP is being replaced in the next 6 months, audit once the new one is live, or the data checks will describe a system you are about to switch off.
• You want a security or compliance audit. We do not run cybersecurity audits, penetration tests or certification readiness work. Our risk notes cover AI-specific risks, such as data sent to model providers and outputs that need human review.

What the audit does not include: building anything beyond feasibility tests, vendor selection RFPs, staff training or ongoing advice. Each of those exists as its own service, and none is required to act on the findings.