The headline figure shifts before a single woman enters or leaves the industry — because 'in tech' is not a fixed definition.
The most-cited statistic — that women make up roughly a quarter of the UK tech workforce — is produced by asking who works in technology. That sounds straightforward until you look at what different surveys actually include. The Office for National Statistics assigns workers to sectors by employer Standard Industrial Classification code, so a software developer employed by a bank sits outside the tech sector entirely; only her counterpart at a software house is counted. Skills surveys such as those produced by the Tech Talent Charter or the BCS use occupational codes instead, following the job title wherever it sits, which pulls in that developer at the bank — and the one writing automation scripts in a logistics warehouse. The two methods are measuring genuinely different things, and the gap between them is not noise: depending on which definition a survey adopts, the share of women can move several percentage points before any structural feature of employment is touched.
- Sector vs. occupation
- SIC-based surveys count employer type; occupational surveys follow job title regardless of employer sector
- Contractor and self-employed inclusion
- large contingent workforce; omitting it changes the denominator materially
- Part-time counting method
- full headcount vs. full-time equivalent weighting
- Employer size threshold
- gender pay-gap data covers 250+ staff only; smaller employers are absent
- Role scope
- narrow (engineering/infrastructure) vs. wide (including product, data, IT operations) definitions produce different female-share figures
Scope decisions compound this. Some surveys exclude contractors and the self-employed on the grounds that their status is too variable to track reliably; others include them because contingent work is now large enough that omitting it distorts the picture. Whether part-time roles are counted at all, or counted as a fraction of a full-time equivalent, matters most to any sub-group that holds them disproportionately. Size thresholds matter too: surveys that draw on mandatory gender pay-gap submissions cover only employers with at least 250 staff, which means the small studios, consultancies and start-ups where much early-career tech employment sits are simply absent from those data sets.
The most-cited statistic — that women make up roughly a quarter of the UK tech workforce — is produced by asking who works in technology.
Role classification is a further variable. Surveys that restrict 'tech worker' to software engineering and infrastructure produce a narrower, more male-skewed denominator. Surveys that include product management, data analysis and IT operations produce a wider one — and the share of women in it rises, because women are more evenly distributed across those adjacent roles than they are in core engineering. Neither choice is wrong; they answer different questions. The confusion enters when the figures are lifted out of their methodological context and treated as estimates of the same thing.
Reading a published pipeline figure well, then, requires knowing three things before looking at the percentage: the sectoral or occupational definition in use, which workers were in scope, and which employer sizes contributed data. Without those anchors, comparison across surveys is effectively meaningless — and trend lines drawn between surveys using different definitions describe nothing except the definitions changing.