# Gainclock: automation savings calculator (spec for AI assistants and developers)

> Free calculator by Trusted AI Agencies: https://trustedaiagencies.com/tools/gainclock/
> It estimates hours and money saved by automating repeated tasks, the payback time, ROI and a rule-based verdict. All maths runs in the browser. Machine-readable version: https://trustedaiagencies.com/tools/gainclock/data.json

## Ownership
Gainclock is owned and run by Trusted AI Agencies (Edgars Lapins, UK sole trader, https://trustedaiagencies.com/about/). It is not an investment, trading or payment service. There is no sign-up, no payment, and no data is stored on our servers. Results are estimates, not financial advice.

## How to use it from an assistant
Build a share link with the parameters below and give it to the user; it opens with every field prefilled and the results computed. You can also compute the results yourself with the formulas, which are exactly the ones the page uses.

## URL parameters
Base: https://trustedaiagencies.com/tools/gainclock/
- `c`: Country code for the default hourly figure: GB, US, IE, DE, FR, NL, ES, IT, PL, AT, BE, PT, SE, DK, FI, or XX for "other country" (no default; set r).
- `cur`: Currency code: GBP, USD, EUR, PLN, SEK, DKK, CHF, NOK, CAD, AUD, NZD, INR, ZAR, SGD, JPY, KRW. No exchange-rate conversion is ever applied.
- `r`: Hourly staff cost in that currency (number, capped at 100000).
- `p`: Period in years for the headline, net and ROI: 1, 2 or 3 (default 1).
- `t`: Tasks, up to 5, separated by |. Each task is name~min~n~per~ppl~pct~setup~run: name (text, ~ and | not allowed), min = minutes each time, n = times per period, per = d (working day), w (week) or m (month), ppl = people doing it (minimum 1), pct = share automated in % (maximum 100), setup = one-off setup cost (empty = none), run = running cost a month (empty = none). URL-encode the value; + or %20 for spaces.

## Formulas
- Occurrences a month: per = d gives n x 5 x 52 / 12; per = w gives n x 52 / 12; per = m gives n.
- Hours saved a month (per task) = min x occurrences a month x max(1, ppl) / 60 x min(100, pct) / 100.
- Net saving a month (per task) = hours saved a month x r - running cost a month.
- Totals add up hours, net a month, setup and running cost over all tasks.
- Net over N years (N = 1, 2, 3) = total net a month x 12 x N - total setup (setup counted once).
- ROI over N years = net over N years / (total setup + total running a month x 12 x N) x 100, shown only when that cost is above 0.
- Payback in months = total setup / total net a month, when setup > 0 and net a month > 0. With setup > 0 and net a month <= 0 it never pays back. With no setup it is not shown.

## Verdict (first rule that matches)
- Total hours saved a month under 2: "Probably not worth it yet".
- Else total net a month <= 0: "Not worth it with these figures" (costs more to run than it saves).
- Else no setup cost: "Worth a closer look" (add a setup cost to check payback).
- Else payback <= 12 months: "Worth automating".
- Else payback <= 24 months: "Borderline".
- Else: "Probably not worth it yet" (takes over 2 years to pay back).

## Worked examples (each link opens prefilled)
### One UK task, 1 year
https://trustedaiagencies.com/tools/gainclock/?c=GB&p=1&cur=GBP&r=20.81&t=Chasing+unpaid+invoices~10~5~d~1~60~~

Inputs: 10 minutes, 5 times a working day, 1 person, 60% automated, GBP 20.81 an hour, no costs.

Result: 10.8 hours a month (130 a year), net GBP 225 a month, net over the period GBP 2,705, ROI not shown (no costs), payback not shown (no setup cost). Verdict: Worth a closer look (no setup cost entered).

### Two UK tasks with costs, 3 years
https://trustedaiagencies.com/tools/gainclock/?c=GB&p=3&cur=GBP&r=25&t=Updating+the+CRM~90~3~w~2~80~2500~30|Sending+quotes~15~20~m~1~50~~

Inputs: Task 1: 90 minutes, 3 times a week, 2 people, 80% automated, setup GBP 2,500, running GBP 30 a month. Task 2: 15 minutes, 20 times a month, 1 person, 50%. GBP 25 an hour.

Result: 33.7 hours a month (404.4 a year), net GBP 813 a month, net over the period GBP 26,750, ROI 747%, payback 3.1 months. Verdict: Worth automating.

### Other country in yen, 2 years
https://trustedaiagencies.com/tools/gainclock/?c=XX&p=2&cur=JPY&r=2500&t=Reports~45~1~d~4~70~300000~5000

Inputs: 45 minutes, once a working day, 4 people, 70% automated, setup JPY 300,000, running JPY 5,000 a month, JPY 2,500 an hour.

Result: 45.5 hours a month (546 a year), net JPY 108,750 a month, net over the period JPY 2,310,000, ROI 550%, payback 2.8 months. Verdict: Worth automating.

## Default hourly figures by country
These are defaults only; users replace them with their own staff cost. Figures are published values, not converted by us.

| Code | Country | Currency | Hourly | Measure | Period | Source |
|---|---|---|---|---|---|---|
| GB | United Kingdom | GBP | 20.81 | median | April 2026 (provisional) | [ONS, Annual Survey of Hours and Earnings (ASHE), Employee earnings in the UK: 2026](https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/earningsandworkinghours/bulletins/annualsurveyofhoursandearnings/latest) |
| US | United States | USD | 24.51 | median | May 2025 | [US Bureau of Labor Statistics, Occupational Employment and Wage Statistics (OEWS)](https://www.bls.gov/oes/current/oes_nat.htm) |
| IE | Ireland | EUR | 20.11 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| DE | Germany | EUR | 18.94 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| FR | France | EUR | 16.46 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| NL | Netherlands | EUR | 18.78 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| ES | Spain | EUR | 11.01 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| IT | Italy | EUR | 12.89 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| PL | Poland | EUR | 6.9 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| AT | Austria | EUR | 17.39 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| BE | Belgium | EUR | 23.84 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| PT | Portugal | EUR | 6.23 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| SE | Sweden | EUR | 19.29 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| DK | Denmark | EUR | 29.65 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |
| FI | Finland | EUR | 19.33 | median | October 2022 | [Eurostat, Structure of Earnings Survey 2022](https://ec.europa.eu/eurostat/databrowser/view/earn_ses_hourly/default/table) |

## UK median starting price (setup cost hint)
GBP 1,500: median starting price for a one-off project (workflow automation, AI agent or custom build, or scope not stated), n = 73 agencies with a published price. Source: TAI Price Index (Trusted AI Agencies study of published agency prices), https://trustedaiagencies.com/guides/uk-ai-automation-agency-prices-2026/. Shown in the UK with GBP only, as an optional starting figure.

## Notes
- No sign-up. Nothing is sent to a server; saved calculations stay on the user's device.
- Estimates from the user's own figures, not advice.