Insights · AI Readiness
What an AI readiness assessment should cost.
The same words cover a free questionnaire and a $150,000 programme. What actually differs between the bands, and how to tell which one your situation needs.
"AI readiness assessment" covers a free ten-minute questionnaire and a fourteen-week programme costing more than most organisations' entire first AI build. Both are sold under the same words, to the same buyers, often in the same week.
The price bands below are what we see in the Australian market in 2026, across mid-market and enterprise buyers. They are ranges, not quotes, and the point is less the numbers than what actually changes between them.
Free: the vendor questionnaire.
Twenty to forty questions, an instant score, a maturity band and a PDF. Offered by platform vendors, by consultancies as a lead magnet, and by us.
What it is good for: structuring a conversation, and finding out quickly whether your leadership team answers the same questions differently. That disagreement is genuinely useful information and costs nothing to obtain.
What it is not good for: deciding where to spend money. The questions are shaped by what the provider sells, they are answered by whoever fills in the form rather than verified against anything, and the output is a position rather than an instruction. Treat it as qualification, not analysis. Ours is the readiness quiz, and it carries the same caveat.
$15,000 to $35,000: the focused assessment.
Two to three weeks, one business area or one candidate use case, two consultants part-time. Six to ten interviews, a hands-on look at the actual data rather than a description of it, and a written output.
This is the band most mid-size Australian organisations need and the one most often skipped in favour of something bigger. What you should get for it:
- A verified view of the data, meaning someone queried it. Row counts, date coverage, duplication rate, where the definitions disagree. Not a description of the systems from an architecture diagram.
- A named first use case, costed. Scope, duration, price band, business owner, and the outcome that changes.
- The disqualifications. Which of the ideas on your list are not viable yet and what specifically would need to change. This is the page that pays for the engagement.
- A short remediation list, sequenced, with the one constraint that binds first called out plainly.
If a quote in this band does not include hands-on data work, it is an interview exercise with a framework on top and should be priced closer to the bottom of the band.
$40,000 to $90,000: the enterprise assessment.
Four to six weeks, several business domains, fifteen to thirty interviews, a regulatory and governance review alongside the data work, and a prioritised portfolio rather than a single use case.
This band is justified when the answer genuinely differs by domain, when you are regulated and the governance review is real work, or when the output has to survive a board paper and a procurement process. For an APRA-regulated entity, the CPS 230 and CPS 234 implications of an AI programme are a material part of the scope, and pricing them into the assessment is cheaper than discovering them mid-build. Our CPS 230 guide covers what that review has to establish.
What should be different at this price, beyond volume: a governance and control assessment that names the specific artefacts you would need to produce, an operating-model recommendation with roles rather than a diagram, and a portfolio sequenced by dependency rather than by expected value. Value-ranked lists are easy and almost always wrong, because the highest-value use case is usually the one with the deepest data dependency.
$120,000 and above: assessment bundled with strategy.
At this point you are buying an AI strategy, an operating-model design and a multi-year roadmap, with readiness as one workstream inside it. That can be the right purchase. It should be bought deliberately, and judged as a strategy engagement, which is a different evaluation.
The failure mode in this band is that the assessment component gets diluted. Thirty interviews across eight domains produce a roadmap that is broad, well-presented, and never specific enough about any one thing to start on Monday. If you buy here, insist that one workstream produces a costed first engagement at the level of detail described in the $15,000 band. The roadmap is worthless without it.
What actually drives the price.
Four things, in order of impact:
- Whether anyone touches the data. Hands-on data profiling is the single largest cost difference between a cheap assessment and a useful one, and it is the line most often absent from a quote.
- Interview count. Roughly linear, and the main driver of duration. Ten interviews is a business area, thirty is an enterprise.
- Regulatory scope. A genuine APRA, Privacy Act or sector-specific review is specialist work and is priced as such. A paragraph noting that obligations exist is not that.
- Team seniority. Whether the people who run the interviews are the people named in the proposal. Ask for the time split by named person, not the blended rate.
Three quotes worth walking away from.
The assessment that cannot price what it recommends. If the output is a score and a recommendation to commission a further scoping exercise, you have paid for a document that argues for another document. An assessment that cannot cost its own recommendation has not finished.
The free assessment from the firm that will do the build. Not because the advice is dishonest, but because the incentive is unmanaged. If the same firm assesses and delivers, pay for the assessment so that "you do not need this yet" remains a commercially survivable answer for them to give.
The fixed six-week assessment with no data access clause. Six weeks assumes access it has not secured. The clause matters more than the timeline, because the timeline will move to match the access.
Which band you need.
If you can already name the decision that would change, the focused band is enough, and the enterprise band is over-buying. If you cannot, or if different parts of the business would name different decisions, the enterprise band is buying you the resolution of that disagreement, which is a legitimate thing to pay for.
If you are not sure which of those describes you, the seven questions in are you AI ready will sort it in an afternoon, free. What the output should then commit you to is covered in what happens after an AI readiness assessment, and the wider pricing picture in what AI consulting costs in Australia.
General information only, not legal or regulatory advice. Current as at September 2026.