Free tool · Calculator
Will this tool pay for itself? The subscription is not the cost.
A £29 tool and a £290 tool are usually compared on price, and the comparison is almost always wrong. Put the hours in and the arithmetic often says no — including for tools that are free.
- Buying AI
- Pricing
- Total cost of ownership
- Getting started
The short answer
Adoption costs hours before it costs money: evaluating, onboarding, and supervising the output for as long as the tool runs. Put your own hours and internal hourly cost in and this returns the first-year adoption cost, the monthly saving that has to be beaten before a subscription is even considered, and whether your expected saving clears it. It can and frequently does answer do not buy this.
On this page · 5 sections
The published version of this argument uses eight hours to choose, twelve to onboard and one a week to supervise, at £30 an hour of internal cost — eighty hours, £2,400, about £200 a month. Those were our numbers and they were published so they could be disagreed with.
These are yours. The supervision figure is the one that decides most outcomes, because it is the only cost that never stops.
Your numbers
Opens on a worked example, not an industry average. Replace every figure with your own — there is no such thing as a default here.
—
Show the working
Why supervision decides it
Move the supervision field to two hours a week and watch the floor. At £30 an hour that one change adds £3,120 a year, which is more than most of the tools being compared cost in total.
That is not an argument against supervision. It is an argument against buying things whose output needs watching and pretending the watching is free — and, for anything client-facing or regulated, the watching is not optional.
What it does not price
Risk, disruption, and the cost of being wrong. A tool that saves four hours a week and produces one bad client letter a year has a cost this does not model. Neither does it price the opposite: a capability you do not have today, which is bought against a different question entirely.
Questions people actually ask
What does it really cost to adopt a new business tool?
The subscription plus the hours: choosing it, onboarding onto it, and supervising its output for as long as it runs. Supervision usually dominates, because it is the only one that never stops — an hour a week at £30 is £1,560 a year, which is more than most tools cost.
Can a free tool still lose money?
Yes, and routinely. If the saving is smaller than the hours it takes to choose, onboard and supervise, the tool loses money at any price including zero. That is why "it is only £29, just try it" is an expensive sentence.
What hourly rate should I use?
Your internal cost — salary plus overhead for whoever actually does the work — rather than your charge-out rate. Charge-out overstates the saving because the hours freed rarely convert one-for-one into billable work.
Does this tool ever say buy it?
Yes, when the saving clears the adoption cost with room to spare. It also says marginal, and it says do not buy this, which is the answer it gives most often at the savings people actually expect.
Sources
- The default hours shown — 8 to choose, 12 to onboard, 1 a week to supervise, at £30 an hour of internal cost — are Noxia’s own model, published in the £200 floor note so they can be disputed. They are a worked example, not a benchmark, and every one is editable here. noxia.co.uk ↗ — our own arithmetic, labelled as such. No survey underlies the default hours
- The point that supervision becomes non-negotiable where output is client-facing or regulated rests on the safeguards in Article 22C of the UK GDPR, substituted by section 80 of the Data (Use and Access) Act 2025 and in force since 5 February 2026. legislation.gov.uk ↗ — primary for the legal point; the costing is ours re-checked every 6 months
- Nothing entered here leaves the page. There is no network call in this file and no field that identifies anybody. — first-party, and enforced: test/pages.js fails the build if this page names any API that could open a connection
Checked 23 September 2026. Next scheduled check 22 March 2027. Numbers that move — leaderboards, live indices — are re-checked every 30 days; annual datasets and rules in force every six months; dated research once a year. If something here has gone stale before we got to it, tell us and we will correct it and say what changed.
If it says no, that is the tool working.
We would rather you did not buy something than bought it and let it lapse. If the arithmetic does clear, tell us which job it is and we will say whether it is worth building properly or whether the off-the-shelf version is fine.
Start the conversation