Trust
Method and data rules
How CercleWork turns official statistics into pages, and what we refuse to publish.
Our rules
- Real data only. Every number comes from a named, dated source, stored in versioned data files. Nothing is estimated or filled in.
- Payslips and job ads are different measures. We label which is which and never blend them.
- Pay figures are shown only where ONS rates them precise or reasonably precise (coefficient of variation 10% or less). Suppressed or weak cells are left out.
- Job advert trends compare the last 12 months with the same months a year earlier, counting only months ONS published in both years.
- A role is called growing or cooling only if adverts moved by 2% or more, and only if it had at least 3,000 adverts. Smaller moves are shown as flat.
- No company league tables, skill rankings or salary pages for places without reliable data. When we can't source a section, we leave it out.
- Reports are generated from the data, then reviewed and edited by a person before publishing.
Refresh
A script downloads the latest ONS files and rebuilds the data. ONS publishes job advert data monthly, vacancies monthly, working arrangements monthly and ASHE pay each October. This site was last refreshed on .
Sources and method
Last updated . Figures are read from the source files and are not estimated.
- ONS Annual Survey of Hours and Earnings (ASHE) 2026 provisional, Table 14.7a, Office for National Statistics, published 22 October 2026. Period: April 2026 (tax year to 5 April 2026). Employer payroll survey (about 1% sample of PAYE jobs). Gross annual pay, full-time employees in the same job for at least a year. Provisional; revised a year later. Accredited official statistics
- ONS ASHE 2025 revised, Table 14.7a, Office for National Statistics, published 22 October 2026. Period: April 2025. Same survey, previous year, revised figures (published with the latest edition). Used only to calculate the year-on-year change. Accredited official statistics
- ONS ASHE 2026 provisional, Table 15.7a (work region by occupation), Office for National Statistics, published 22 October 2026. Period: April 2026. Regional medians by 4-digit SOC 2020. CercleWork shows only cells with a coefficient of variation of 10% or less (precise or reasonably precise). Accredited official statistics
-
ONS Labour demand volumes by SOC 2020 (Textkernel online job adverts), Office for National Statistics, published 23 October 2026. Period: January 2017 to September 2026.
New online job adverts scraped from about 90,000 job boards, de-duplicated by ONS. Counts adverts, not vacancies. Official statistics in development. Official statistics in development Method guide.
ONS data-quality notes
- Since November 2025, the data we have been publishing has shown increased uncertainty, and so users should exercise more caution in tracking month on month changes, and where possible, instead consider the longer term trends. We are looking at identifying ways to deal with this uncertainty in the coming few months.
- In the October 2026 publication, we have partially imputed the data from October 2025 to February 2026 and July to September 2026. This is due to issues with a major Textkernel source, from which we were unable to obtain accurate information for this period. These issues included a high volume of duplicate adverts and Textkernel being unable to scrape the website. The adjustment applied to the data has been calculated based on the average trend of the source for the past year before the issue arose, and it has been applied consistently across every table. Hence, users should treat with caution any individual odd month-on-month trends during those periods, which are more likely to be the case for more granular occupation and geography series, and focus more on the longer term trends. Please note, some of the tables may now not sum up due to rounding.
- The data in March 2026 has been suppressed due to issues encountered when obtaining job advert data from a major source. We are aiming to recover this data and may include March 2026 figures in future outputs, and we will update users accordingly.
- Users should be aware that a large source has increased its level in September 2026 to what has correctly been identified as the correct level. This may result in significant increases between August and September 2026.
- ONS VACS02: Vacancies by industry (Vacancy Survey), Office for National Statistics, published 17 November 2026. Period: Aug-Oct 2026. Survey of businesses. Rolling three-month averages, seasonally adjusted, thousands. Counts open vacancies by industry, not occupation. Accredited official statistics
- ONS Opinions and Lifestyle Survey: working arrangements, Office for National Statistics, published 16 October 2026. Period: 2 to 27 September 2026. Household survey of adults in Great Britain. Share of working adults who, in the past seven days, worked from home only, travelled to work only, or did both (hybrid). All working adults, not tech workers only. Official statistics
- CV-Library/APSCo, UK tech jobs market Q2 2026: AI hiring accelerates, CV-Library and APSCo, published 12 August 2026. Period: April to June 2026 (Q2 2026). Job adverts posted on CV-Library. Advertised pay, not payslips.
- PwC UK AI Jobs Barometer 2026, PwC UK, published 15 June 2026. Period: Calendar year 2025. Analysis of more than one billion job adverts in 27 countries and territories; UK figures. Advertised pay in job ads, not payslips.
- KPMG and REC, UK Report on Jobs, November 2026, KPMG and the Recruitment & Employment Confederation (compiled by S&P Global), published 9 November 2026. Period: October 2026 (survey 12 to 26 October). Monthly survey of about 400 UK recruitment consultancies, reported as diffusion indices. Direction only, no counts or pay.
- Hays UK Salary & Recruiting Trends 2026, Technology guide, Hays plc, published December 2025. Period: Data collected during 2025. Hays UK offices' job listings, offers and candidate registrations in 2025 (method on p.24). Typical salaries and ranges, not medians. No sample size per role.
How we read the data: method and data rules.