Six engagements. One outcome: AI that actually pays for itself.
A clear-eyed audit. A stack that fits. Workflows your team uses. A plan to keep both the regulator and the CFO happy. Each service is fixed-fee, scoped before kickoff, and ends with a working deliverable, not a deck.
AI Readiness Audit
A 2-week structured diagnostic. Where you are, where AI can plausibly add value, what's noise, and what to fix first. The output is a board-ready document, not a sales pitch.
What's actually included
- Stakeholder interviews (typically 6-10 people across leadership, operations, and IT)
- Current-state stack assessment, mapped against your real workflows
- Data hygiene check (the unglamorous bottleneck most audits skip)
- Workflow-by-workflow ROI mapping, scored on impact and effort
- Risk and compliance register (Privacy Act, sector rules, data residency)
- 12-month prioritised roadmap with sequencing and rough cost ranges
Who it's for
Firms with 10-200 staff, already using some AI tools (or thinking about it), where leadership wants a clear yes/no/not-yet on which AI investments are worth making in the next 12 months.
Who it's not for
Firms looking for an outsourced AI strategy to sign off without internal involvement. The audit needs your people in the room. If they're not available for interviews, the audit can't do its job.
Sample deliverable
A 40-50 page audit document with executive summary, ranked opportunity list, risk register, and quarterly roadmap. Plus a 90-minute board walkthrough.
Tool Stack Selection
There are 600+ AI tools claiming to do your job. Most don't. This service is the structured way to choose, with real testing and total-cost-of-ownership modelling, instead of vendor pitch theatre.
What's actually included
- Requirements gathering with the people who'll actually use the tools
- Long-list build (typically 15-30 vendors), filtered to a short-list of 3-5
- Side-by-side feature, security, and pricing comparison
- Live pilot of top 3 candidates inside your environment with real users
- Total cost of ownership over 3 years, including switching costs
- Procurement support and contract redline
Who it's for
Firms about to commit $20k-$300k+ on AI tooling who want vendor-independent advice before signing. Especially common after a year of trialling tools that didn't stick.
Who it's not for
If the decision is already made and the brief is "validate our pick," we'll politely decline. We're not a rubber stamp.
Sample deliverable
A ranked recommendation memo with scoring rubric, pilot results, 3-year TCO, contract red-flags, and a one-page board summary.
Workflow Integration
Not slides about AI. A working workflow your team uses on Monday. We embed with your team for 4-12 weeks, build it, train them, and hand it over running in production with documentation.
What's actually included
- Workflow design with the team who'll use it (not for them)
- Prompt engineering, system prompts, and guardrails
- Integration with existing systems (CRM, document store, email, etc.)
- Verification step design (so output is checked, not blindly trusted)
- Team training, with structured practice in the first 4 weeks
- Production handover, full runbook, and 30-day post-go-live support
Who it's for
Firms who've completed an audit (with us or otherwise) and are ready to ship a specific high-ROI workflow. We expect the audit work to be done so we can spend the engagement building, not deliberating.
Who it's not for
Firms still working out whether AI is right for them. Start with the Readiness Audit first. Building before you've diagnosed is the expensive mistake we exist to prevent.
Sample deliverable
A live, in-production workflow with prompt library, runbook, training videos, adoption dashboard, and a named owner inside your team.
Risk + Compliance Review
Especially relevant for regulated industries (financial advice, accounting, legal, health). FMA, FAP licensing, IRD, Privacy Act 2020, copyright, data residency. We map your AI use against the rules and tell you exactly what's safe.
What's actually included
- Privacy Act 2020 assessment (lawful basis, retention, breach response)
- Sector-specific compliance review (FMA, MoJ, IRD, Medical Council)
- Vendor data-residency and sub-processor review
- Disclosure obligations and client-facing language
- Acceptable use policy drafted and ready for board sign-off
- Incident response protocol for AI-specific failure modes
Who it's for
FAP-licensed financial advisers, accounting partners, lawyers, and health practices using or planning to use AI in any client-facing capacity. Also boards who want to demonstrate AI governance to regulators or auditors.
Who it's not for
Firms with zero current AI use and no near-term plans. The review is high-leverage when there's something concrete to assess, not as an abstract exercise.
Sample deliverable
A risk register mapped to the relevant NZ legislation, a board-ready compliance memo, a finished acceptable use policy, and an AI incident-response runbook.
Fractional AI Director
4-6 hours a month, ongoing. Strategic oversight without hiring a full-time head of AI. The role most NZ SMBs need but can't justify hiring for at $200k+ a year.
What's actually included
- Monthly strategy review with leadership (60-90 min)
- Vendor pitch sit-ins, with our notes and recommendation after
- Quarterly stack health check and re-rank against new tooling
- Quarterly board memo on AI investment status and next moves
- On-call sanity-check before signing any new AI contract
- Annual roadmap refresh as part of strategic planning
Who it's for
Firms past the initial implementation phase who need ongoing strategic oversight but don't yet have the volume to justify a full-time AI lead. Usually 20-150 staff, with AI now embedded in 2-5 workflows.
Who it's not for
Firms looking for hands-on workflow building. The fractional role is governance and oversight, not delivery. For build work, see Workflow Integration.
Sample deliverable
Quarterly: a 6-page board memo (AI status, ROI to date, vendor moves, recommended next steps). Monthly: meeting notes and an updated stack scorecard.
Team Training
Practical workshops, not theoretical AI. Here's how you use Claude, ChatGPT, Copilot, or whatever else makes sense, for your actual day job. One day on-site, then four weeks of structured practice.
What's actually included
- Role-specific use cases (we don't teach accountants the same things we teach paralegals)
- Hands-on day with real client work, not toy examples
- A team-specific prompt patterns library, taken home as a working document
- Four weeks of structured practice with weekly group check-ins
- Measurable adoption metrics (who's using it, on what, with what result)
- Safe-use guidance (data confidentiality, verification, escalation paths)
Who it's for
Teams of 6-40 already given AI tool access but underusing them. Common pattern: subscriptions paid, three power users active, everyone else avoiding. We fix the everyone-else part.
Who it's not for
Generic AI literacy training across a 500-person org. We do focused team uplift, not enterprise-wide e-learning rollouts. For that, we'll refer you to a partner who specialises in it.
Sample deliverable
A workshop day, a working prompt library tailored to the team's actual work, four practice check-ins with metrics, and a final adoption report.
Four phases. No deck-only deliverables.
Most AI consulting engagements end with a slide deck and a hopeful look. We end with a workflow your team uses, with documentation, with measurable ROI, and a plan for what comes next.
Diagnose
Understand the business, the existing stack, the team's capability, the constraints. Output: structured audit, risk register, top-10 ROI candidates.
Roadmap
Prioritise the candidates by impact and effort. Sequence them. Choose the tools. Get sign-off from the people who'll actually use it.
Implement
Build the first workflow. Embed with the team. Iterate. Measure. Ship to production once it's doing what we promised.
Sustain
Models change. Vendors disappear. Compliance rules tighten. We can stay on as fractional AI director, or hand it cleanly to your team with documentation.
The most expensive AI mistake isn't picking the wrong tool. It's building before you've diagnosed. Audit first, build second.Craig Coupland · Founder, Coupland Consulting
What firms ask before they start.
How is this different from a regular IT or strategy consultant?
Most IT consultants stop at "here are the tools, good luck." Most strategy consultants stop at "here's a slide deck, good luck." We do both ends and the middle: clear audit, tools that fit, working implementation, and ongoing review. The deliverable is a workflow your team uses on Monday, not a deck the executive team reads once.
What does an engagement actually look like?
Most engagements start with a 30-minute call to scope the problem. From there, the typical sequence is a 2-week Readiness Audit, then 1-2 weeks of tool selection, then 4-12 weeks of workflow implementation. Some clients then move onto a fractional AI Director monthly retainer. Some hand it back to their internal team with our runbook. Both are fine.
What's the typical cost?
An end-to-end engagement (audit + tool selection + first workflow deployment) usually lands between $25k and $60k+GST depending on scope and team size. Ongoing fractional AI Director is $2.8k+GST/month for a 4-hour commitment. Standalone services start at $4.5k+GST. All quotes are fixed-fee, agreed in writing before kickoff, with no hourly billing or scope-creep invoicing.
Do you work with regulated industries?
Yes, and this is where we add the most value. Financial advice (FAP-licensed firms), accounting, legal, and health all have specific rules about AI use, data residency, disclosure, and supervision. Craig's CFP designation and FAP-licensed firm background means we speak that language fluently. We've worked through Privacy Act 2020, FMA expectations, IRD positions on automated reasoning in tax, and Medical Council guidance on AI in clinical settings.
What if we're not ready for AI yet?
Then we tell you. Plenty of firms aren't ready. Sometimes the bottleneck is data hygiene, sometimes team capability, sometimes leadership alignment, sometimes the simple fact that the business problem doesn't actually need AI to solve it. The Readiness Audit will tell you specifically what to fix first, in what order, before AI is even worth attempting. That conversation alone saves clients tens of thousands.
Are you tool-agnostic or do you push specific vendors?
We take no commissions, referral fees, or kickbacks from AI vendors. Ever. Our recommendations are based entirely on real testing inside your stack and with your team. If the free or low-cost tool wins the pilot, we recommend the free tool. If the right answer is "build nothing yet, switch to a better notes app," we tell you that. Vendor-neutrality is the entire reason this practice exists.
How do you handle NDAs and data confidentiality?
Standard NDA on every engagement, signed before scoping calls. For regulated clients we agree data handling protocols upfront in writing: which model providers may or may not be used, where any data may sit, what gets logged, what gets retained, and what gets deleted post-engagement. We never use client data to demo other clients, ever.
What success metrics do you actually track?
Per workflow: time reclaimed per week, error rate vs. baseline, percentage of the team using it weekly, dollar cost per task. Per engagement: payback period and 6-month ROI. We do not count slides delivered, hours billed, or meetings attended. If a workflow isn't paying for itself within an agreed window, we tell you and we change course. That's the deal.
AI tools change every quarter. How do you handle that?
We choose tools for the workflow they unlock, not the model behind them. Our handover runbook deliberately includes how to swap the underlying model without rebuilding the workflow, so when GPT-7 or Claude 6 lands you're not back at square one. Fractional retainer clients get a quarterly stack review built in. Workflow Integration clients get a 30-day post-go-live window where we'll absorb any vendor-side surprise that affects the build.
Is Coupland Consulting separate from Wealth Health?
Yes. Craig founded Wealth Health (Tauranga financial advisory firm) in 2014 and still runs it. Coupland Consulting is a separate practice, separate brand, focused on AI strategy for businesses. Same person, same standards, different scope. AI consulting isn't financial advice, so there's no licensing overlap. We just keep the brands separate so nobody's confused about which hat we're wearing.
Want to know which service your firm actually needs?
30-minute discovery call. No deck, no pitch. Tell us about your business and we'll tell you, honestly, which service makes sense (or whether to wait six months).