Before you buy AI, find out if your data can support it.

An AI readiness assessment is a structured review of whether an organization can actually get value from AI yet. It looks at four things: the data (is it accessible, accurate, and complete enough to feed a model or an agent), the processes (which tasks are repetitive and rule-based enough to automate), the people (who would use the output, and would they trust it), and the tooling (what systems exist, and can they be connected). The output is a readiness score per function and a ranked list of opportunities, so you spend on the ones that will work rather than the ones that demo well.

We score each function against the same four pillars, using evidence rather than opinion. Data readiness is measured by looking at the actual systems: what is captured, where it lives, how clean it is, and whether it reconciles. Process readiness comes from mapping how work is really done, including the manual steps nobody documents. People readiness is assessed through interviews with the people doing the work. Tooling readiness is a review of what you already pay for and what those tools can already do. Each function ends up with a score and a short explanation of what would have to change to raise it.

Data, process, people, and tooling. Data is whether the information a model would need exists and can be trusted. Process is whether the work is structured enough to hand over. People is whether the team would adopt and verify the output rather than quietly ignore it. Tooling is whether your existing stack can be connected without a rebuild. Most failed AI projects fail on data or people rather than on the model itself, which is why an assessment that only looks at use cases tends to produce a list of things that never ship.

Completely different, and the naming causes real confusion. An AI compliance audit examines AI systems you already run, checking them for bias, transparency, and conformity with frameworks like ISO 42001 or the EU AI Act. That is a regulated assurance service. This is the opposite direction: you have little or no AI in place, and the question is where it would help and whether your foundations can support it. If you need conformity assessment for a deployed model, we are not the right firm and we will say so.

It is a fixed-scope engagement quoted after a discovery call, because the price depends on how many functions are in scope and how many systems have to be reviewed. A single-function assessment is a much smaller piece of work than a company-wide one. Market rates for comparable readiness and maturity assessments generally sit in the low-to-mid four figures for a focused scope and higher for enterprise-wide reviews. We scope it to the smallest version that would still change a decision.

Whichever ones carry the manual load. For ecommerce teams that is usually customer support, merchandising and content production, marketing operations, reporting, and finance or operations admin. We start from where your team actually loses hours rather than from a fixed template, so a brand drowning in support tickets gets a different assessment from one whose bottleneck is producing product content at scale.

The engagement includes the scored framework and the completed assessment for your business, which you keep regardless of whether you work with us afterwards. If you would rather run a first pass yourself, ask us on a call and we will share the checklist we use, because a team that has already thought about its own processes gets far more out of the full assessment.

You get a ranked opportunity list with an estimate of the manual hours each one removes and what would have to be true for it to work. Some items will be things your existing tools already do and nobody switched on. Some will need data work first, which is where our tracking and reporting practice comes in. Some will need building. You can take the roadmap to your own team, to another firm, or to us, and the assessment is written to be useful in all three cases.