Pricing
What this costs.
Priced per project, never per month. Nothing here is payable before we have spoken and agreed what is in scope — the first conversation costs nothing and often ends with me saying you do not need it.
Most people start here
Discovery & Roadmap
Find out where AI actually pays off in your organisation before committing budget to it.
- AI readiness assessment across five parameters
- Opportunity mapping: 5–7 concrete use cases
- Strategic roadmap for the next six months
- Pilot plan with ROI calculations
From $10,000
2–4 weeks
You are left with
A costed six-month roadmap with the use cases ranked by return.
Right for
Anyone who has been asked to “do something with AI” and wants to know what is actually worth doing before spending on it.
You are buying a decision, not a commitment. Plenty of clients take the roadmap and run the rest themselves — that is a perfectly good outcome and I will say so if it is the right one.
What moves the number
The figures below are starting points, not rate cards. Four things decide where a project actually lands, and you control two of them.
- How many use cases
- One well-chosen use case costs a fraction of five. Discovery exists partly to work out how few you actually need.
- The state of your data
- This is the single biggest variable. Clean, accessible data can halve a timeline; data nobody has touched in years can double it.
- How much your team carries
- If you have engineers who can build, I design and steer and the cost drops. If you need delivery as well, it rises.
- Regulatory load
- Work that has to satisfy an auditor carries documentation and review that unregulated work does not.
Where it goes from there
Only relevant once the roadmap exists and you have decided something is worth building. Read these as directions, not as a menu you order from today.
Pilot & Rollout
2–6 months
Turn one or two of those use cases into working software your team can operate.
Teams that already know which use case they want and need it built and adopted, not prototyped again.
From $35,000Enterprise AI Programme
6–12 months
A multi-year programme when AI has to work across departments, not in one corner.
Organisations running AI in more than one department, where the problem is now governance and consistency rather than a first win.
From $200,000ISO 9001 from zero
12–16 weeks
A quality management system that passes audit without turning the company into a paperwork factory.
Outsourcing and product companies losing shortlist places for want of a certificate, with no appetite for a year-long programme or a full-time quality hire.
From $25,000Delivery & PMO
Project-based
The programme management underneath all of the above — available on its own.
Programmes in trouble, or teams that need delivery structure without an AI component at all.
Custom
How this compares
Worth knowing what the same work costs elsewhere, so the numbers above have something to sit against.
Published hourly rates by firm type, US dollars
| Firm type | Hourly rate range |
|---|---|
| Independent consultants (where I sit) | $150 to $350 per hour |
| Boutique AI firms | $350 to $650 per hour |
| Accenture, IBM, Capgemini | $400 to $700 per hour |
| Deloitte, PwC, EY, KPMG | $400 to $800 per hour |
| McKinsey, BCG, Bain | $500 to $1000 and above per hour |
Assessment
Published ranges put a mid-market AI readiness assessment at $5,000–$15,000 and an enterprise one at $15,000–$50,000. Mine starts at $10,000.
Pilot
A proof of concept is commonly quoted at $50,000–$250,000. Mine starts at $35,000, because there is no bench to keep busy.
Programme
A full Big-4 programme typically starts near $500,000 and often passes $1M; a senior-led boutique runs $150,000–$400,000. Mine starts at $200,000.
The gap is not a discount and it is not a different quality of engineering. It is the absence of a pyramid: you are buying the person who does the work, without the partner, the engagement manager and the two analysts learning your business at your expense. What the large firms genuinely sell on top is organisational cover. If that is what you need, buy it — I will tell you when it is.
What people ask about price.
- Where do most engagements start?
- Most start with a Discovery & Roadmap engagement: two to four weeks, from $10,000, ending in a costed six-month roadmap with the use cases ranked by return. It exists so that nobody commits a large budget to an AI programme before knowing which parts of it pay for themselves. It is deliberately small enough to be worth buying on its own, and it does not oblige you to buy anything after it.
- What does an engagement cost?
- Engagements run from $10,000 for a two-to-four-week discovery, $25,000 for an ISO 9001 quality system, $35,000 for a pilot taken into production, and from $200,000 for a multi-year enterprise programme. Pricing is per project rather than a monthly subscription or a seat licence. Four things move the number: how many use cases are in scope, the state of your data, how much of the build your own team carries, and whether the work has to satisfy an auditor.
- Can you work with an in-house team, or do you bring your own?
- Both, and the difference shows up in the price. If you have engineers who can build, the work is design, sequencing and steering, and it costs less. If the build has to be delivered as well as designed, that is a larger engagement. Either way the aim is the same: the team that will operate the system is involved while it is being built, not handed it at the end.
- How long does ISO 9001 certification take from nothing?
- Twelve to sixteen weeks is a realistic window for a company starting with no quality management system, and the engagement starts at $25,000. That is not a theoretical estimate: one remote-first software outsourcing company went from a blank page to a passed certification audit in twelve weeks, with 37 quality artefacts written and approved by six people at a peak load of 0.3 FTE each — no dedicated quality hire. The long version of that project is on this site as a case study.
- What happens at the end of an engagement?
- Your team takes ownership, which is the intended ending rather than a wind-down. Handover is part of the work from the start: the people who will run the system take part in building it, the decisions behind it are written down, and the engagement finishes when the team no longer needs the consultant. An engagement that leaves an organisation dependent on an outside expert has failed at the thing it was bought for.
Want a number for your case?
Thirty minutes is usually enough for me to tell you which engagement fits and roughly what it costs. It is also enough for me to tell you that you do not need one, which happens often enough to be worth mentioning.