Pricing
Published prices, and the spend below which we tell you not to bother.
Fixed prices against published dates. The ceiling we hold ourselves to is on this page, and so is the point at which the arithmetic stops working — which is the number most firms leave out.
Build
One price per offer, against the date on its page.
Which offer fits is a question about scope, not about your budget. We agree it in writing before we start and it does not change mid-build — if the scope we agreed turns out bigger than we thought, that is our estimating error and our cost. What changes the price is you changing the scope, and then we requote before doing the work rather than after.
| Offer | Date | Price | We'd decline below |
|---|---|---|---|
| The First Win | 2 weeks | $12,000 | $1,000/mo |
| Launch Sprint | 4 weeks | $26,000 | Runway, not payback |
| Tool Replacement | 6 weeks | $36,000 | $3,000/mo |
| Tool Replacement · with migration | 8 weeks | $46,000 | $3,833/mo |
| Platform Replacement | 10 weeks to first system | $90,000 | $7,500/mo |
Two of these are priced on something other than duration, and we would rather say so than have it inferred. The First Win is priced at what it costs us to deliver, because it is how we earn the work rather than how we make money on it. Platform Replacement is priced on complexity — enterprise integration, several stakeholders, a compliance surface — and it buys the first system only, never a total. A total would be a guess, so we cost the roadmap afterwards, one phase at a time, in payback order.
These are prices per offer, not a weekly rate — dividing one by its duration will not predict another. Launch Sprint costs more per week than a replacement does, for two reasons: it publishes a hard date, and a missed date means week five onward is on us until it ships. And a business that has not launched has no subscription being retired to measure the price against.
Diagnosis
Paid, credited in full, and worth it on its own.
Priced as expertise rather than as a share of the build it precedes, because reading your actual invoices and telling you which three of twenty tools are worth touching is the part that is hard.
| Offer | Duration | Price | Mechanism |
|---|---|---|---|
| Verified Teardown | 5 working days | $9,500 | Credited in full · 90 days · one per client |
| Cloud Exit Assessment | 2–3 weeks | $20,000–25,000 | Advisory only |
The credit is in full against any build that follows, so if you go on to engage us the diagnosis cost you nothing in the end — and if you do not, you paid for five days of work and got five days of work.
We do not sell it as a place in a queue. We take three starts a month — a First Win or a build, counted the same — not because we could not take a fourth, but because every engagement is reviewed by one person and the fourth is where that attention goes thin. If the month is full we will tell you which month is not, and there is no way to pay your way forward.
After the build
Three options, and the first one costs nothing.
90 days of defect liability is included in every build: fixes against the acceptance criteria, no charge. After that you choose, and the honest default is the cheapest one — most clients with an engineer should run it themselves, and we would rather say so than sell a retainer you do not need.
| Run it yourself | AMC | Operated | |
|---|---|---|---|
| Fixes against the accepted acceptance criteria | — | ✓ | ✓ |
| Security and dependency patches | — | ✓ | ✓ |
| Compatibility updates and API deprecations | — | ✓ | ✓ |
| An agreed change allowance | — | ✓ | ✓ |
| Business-hours support | — | ✓ | ✓ |
| Monitoring and alerting | — | — | ✓ |
| Incident response and an on-call rota | — | — | ✓ |
| Restores tested rather than assumed | — | — | ✓ |
| An availability commitment | — | — | ✓ |
Run it yourself. You own the repository and the runbook, and we hand over a dependency policy and a witnessed restore test at the end of the build. If you have an engineer, this is usually the right answer.
AMC — 15% of the build a year, floor $3,600/yr. A share of build, because that is what maintenance actually tracks. The floor binds on small builds: administering the contract costs the same whatever the build did.
| Build | Build price | AMC / year | Per month |
|---|---|---|---|
| The First Win | $12,000 | $3,600 (floor) | $300/mo |
| Launch Sprint | $26,000 | $3,900 | $325/mo |
| Tool Replacement | $36,000 | $5,400 | $450/mo |
| Tool Replacement · with migration | $46,000 | $6,900 | $575/mo |
| Platform Replacement | $90,000 | $13,500 | $1,125/mo |
Operated is priced on what we commit to — cover and response time — not as a percentage of the build. A share of build is the AMC benchmark above; it does not fund a pager.
| Cover | Response | Monthly |
|---|---|---|
| Business hours | Next business day | $2,400/mo floor |
| Extended hours, with change allowance | 4 hours | $4,200/mo |
| 24/7 on-call | 1 hour | Quoted a rota, not a person — not offered below 4 systems under management |
Cancel any month, with the exit handover written in from the start rather than negotiated once you are already unhappy. Hosting stays in your own cloud account, on your own bill — these figures are our labour, not your infrastructure.
Longer terms
Commit for longer, pay less.
Published as a schedule, not settled in a call. Every figure is a rule we apply, not a concession we make — and we never show a struck-through price beside a lower one. If a number here were negotiable it would not be a published price.
| When | Off | Note |
|---|---|---|
| Second and third system under one agreed roadmap | 10% | and the qualification floor falls with it, so a committed programme reaches tools a standalone replacement cannot |
| Fourth system onward | 15% | |
| Twelve-month commitment instead of rolling, on anything monthly | 15% | |
| Paid in full upfront | 3% | not on the First Win — post-paid is that offer, and it is priced at cost |
Stacked, no engagement is discounted by more than 20%. The parts compound past that, and a schedule that can be arithmetic’d down to nothing is not a schedule.
Where you cannot replace
Cloud Cost Optimisation — 25% of what we save you.
A share of first-year savings only, measured before and after, never ongoing. A permanent cut of savings on infrastructure you own would contradict the whole argument we are making. Below roughly $27,000/mo of cloud spend the assessment costs more than it recovers, and we will tell you that rather than sell it.
The invoicing schedule and the measurement dispute process are still being settled, so this one is agreed in writing case by case rather than published here.