Skip to content

Know which AI is worth funding before you fund it.

A fixed-scope assessment of your processes, systems and data. It ends with a ranked, costed plan: the two or three automations most likely to pay back, what each would take to run safely, and the order to build them in.

Start hereEffort and risk to deliverValue to the businessIllustrativeMonth-end reportingSupplier invoicesCRM clean-upBoard packForecasting modelCustomer onboardingContract reviewChatbot on the website
Fixed scopeAgreed in writing before we start.
Fixed feeQuoted after the discovery session. No day rates, no open-ended extensions.
Weeks, not monthsThe timetable is set at the start and kept.
Led by TomTom Buckley runs every assessment personally.

Who it is for

Chief executives, finance and operations leaders of mid-sized organisations across Australia and New Zealand, usually in one of three situations.

You have been told AI could help but nobody has shown you where, what it would cost or how it would be controlled.

You have run a pilot or a vendor demo that impressed, and it has not made it into how the business runs.

Your board is asking what the organisation's position on AI is, and you want an answer grounded in your own processes.

Not sure whether a process is a candidate? Score one yourself in six questions.

How it runs

Five stages, in order. Your people spend a few hours each in interviews; the rest is our work.

  1. 1 DiscoveryShort interviews with leaders and with the people who do the work, about where time goes and where errors cost money.
  2. 2 Systems and dataA look at the systems you run, how they connect, what access is possible and how clean the data really is.
  3. 3 Opportunity registerEvery candidate process scored on value, effort and delivery risk, so the shortlist is a decision rather than a guess.
  4. 4 Business casesThe top two or three costed properly: today's baseline, the target, the build, the running cost and the controls.
  5. 5 ReadoutA working session with your executive team or board, and a plan you can approve, fund or put to tender.

What you receive

Everything is yours to keep, whether or not you build with us.

  • Opportunity register Every use case we found, scored on value, effort and risk, with the reasoning written down so you can challenge it.
  • Two or three business cases Baseline, target, build cost, monthly running cost and payback for the automations most likely to return the investment.
  • Systems and data readiness What each shortlisted use case needs to connect to, what access it needs and what has to be fixed in the data first.
  • Controls plan Who is accountable, what the agent may do alone, what needs approval and how it is monitored, mapped to Australia's Guidance for AI Adoption and, for New Zealand, MBIE's Responsible AI Guidance.
  • A sequenced roadmap The order to build in, what each stage proves, and the decision points where you can stop or continue.
  • A board-ready summary A short paper your executive team can take to the board, in plain English, with the risks stated.

What the documents look like.

Two sample pages, so you can see the level of detail before you commit to anything. First, the opportunity register.

  1. Every candidate countedThe register lists everything we found, including what we recommend you leave alone.
  2. The reasoning, in writingEach score comes with the evidence behind it, so you can challenge it.
  3. One method for every processValue, effort and risk are scored the same way throughout, so the ranking holds up.
AI Opportunity AssessmentSample page

Opportunity register

Processes reviewed23across four business areas
Shortlisted5scored and ranked
Business cases3costed in full

The shortlist

ProcessValueEffortRiskRank
Supplier invoice processingHighLowLow1
Month-end management reportHighMediumLow2
Customer enquiry triageMediumLowMedium3
Stock reorder suggestionsMediumMediumMedium4
Contract reviewMediumHighHighLater

Why invoice processing ranks first

  • About 1,400 invoices a month are keyed by hand, in two systems.
  • The data is ready: most arrive as PDFs from the same 30 suppliers.
  • A person keeps approving every payment, so the delivery risk is low.
Scored on value, effort and delivery riskReasoning written down for every score

Illustrative. Names and figures are invented.

And the summary for your board.

The paper your executive team can table as it stands: the recommendation, the risks, and the points where the board decides again.

  1. The numbers a board asks forBuild cost, running cost and the return, each measured against today's baseline.
  2. Risks stated, not buriedWhat could go wrong, how likely it is and what the plan does about it.
  3. Places to stopEach stage ends at a decision point, so the board funds one step at a time.
Board paperFor decision

AI: where we invest first

Recommended first buildInvoicesranked 1 of 5
Running cost$2,100a month, all five cost lines
Staff time returned110 hoursa month, against baseline

Sequence and decision points

StageWhat it provesBoard decides
1. Invoice processingAgents can run inside finance, with approvalsWeek 8
2. Management reportReporting can be trusted without re-keyingWeek 16
3. Enquiry triageCustomer-facing drafts meet your standardWeek 24

The recommendation

  • Fund the first build now. Decide on the second after eight weeks of measured results.
  • Risks: data quality in the supplier file, and reliance on one model provider. Each has a mitigation in the plan.
  • Controls: payments above the approval limit still need a person, and every action is logged.
Mapped to Australia's Guidance for AI AdoptionWritten in plain English, with the risks stated

Illustrative. Names and figures are invented.

Download the sample report

PDF, seven pages. A cut-down version of every document you receive, written for a company that doesn’t exist. The fee and the build cost are left out.

What a build costs to run

Each business case sets out the monthly running cost before you commit to a build. The model's usage is one line of that cost, and you can estimate it here. It is separate from the fee for the assessment, which is below.

Estimate the model cost of your own task

Move the sliders to match a task in your business: how often it runs, how much the agent has to read each time, and how much it writes back.

ModelClaude Sonnet 5.5
Per taskUS$0.16
Per monthUS$160
About US$1,920 a year. Model usage only, in US dollars, at Anthropic's published prices read on 10 November 2026. Hosting, monitoring, human review and maintenance come on top. Illustrative: measure your own volumes in a pilot.

The other four cost lines, and how to keep each one down, are in the briefing What an AI agent really costs to run.

What it costs

A fixed fee, agreed in writing before any work starts. It depends on the size of the organisation and how many business areas are in scope, so Tom quotes it after a free thirty-minute discovery session rather than guessing here.

There are no day rates and no open-ended extensions. If the scope changes, the fee changes only with your written agreement.

Book a free discovery session

Built to run in production, like everything we make

  • Inside your systems Agents work in the CRM, finance and operations tools you already run, with the same permissions as a member of staff.
  • A person approves what matters Payments, customer messages and anything above a limit you set wait for a human yes.
  • Every action logged Each step is recorded, so you can answer what happened, when and why.
  • Measured and handed over Results are measured against today's baseline, and your team is trained to run and change it.

Questions people ask us

What is an AI opportunity assessment?

An AI opportunity assessment is a fixed-scope review of an organisation's processes, systems and data that identifies where AI agents and automation would pay back. Spurbuck Advisory's assessment scores every candidate on value, effort and delivery risk and ends with costed business cases, a controls plan and a sequenced roadmap.

How much does the AI Opportunity Assessment cost?

It is a fixed fee agreed in writing before work starts. The fee depends on the size of the organisation and how many business areas are in scope, so Tom Buckley quotes it after a free 30-minute discovery session. There are no day rates or open-ended extensions.

Do we have to use Spurbuck Advisory to build what the assessment recommends?

No. The opportunity register, business cases, controls plan and roadmap are yours to keep and can be used to build in-house, put to tender or approve with Spurbuck Advisory.

Does the assessment follow Australian and New Zealand AI guidance?

Yes. The controls plan maps each recommended automation to the six essential practices in Australia's Guidance for AI Adoption and, for New Zealand organisations, to MBIE's Responsible AI Guidance for Businesses.

Start with a free discovery session.

Thirty minutes with Tom to talk through your systems, where the time goes and whether an assessment is the right next step. You leave with a clear view of what is worth automating, what it would take, and what it would return. No obligation.

We use the details you give us only to arrange and hold your session. See our privacy policy.