We use aggregated and anonymised LinkedIn data to uncover new information on skills supply. LinkedIn has over 23 million members in the UK, just under 10% (2.2 million) of these members have tech skills. We use data which describes the skill profiles of these members, based on the information that members enter on their profiles. The data includes information on:

• Industry Members indicate their current and previous employers in the experience section of their profile. The industry in which a member works is determined by the classification of the company in LinkedIn’s taxonomy of industries.
• Function Members indicate their current and previous job titles in the experience section of their profile. The function in which a member works is determined by the classification of the job title in LinkedIn’s taxonomy of functions.
• Location LinkedIn determines a member’s location by the location they have indicated in their profile summary. When this location is changed, LinkedIn measures that change as migration.
• Education Members indicate their academic achievements in the education section of their profile, such as their higher education organisation and degree type.
• Skills Members indicate their expertise within the skills section of their profile. LinkedIn clusters the tens of thousands of individual skills that members choose to display on their profile into categories for analysis. Each skill cluster is then further aggregated into ‘parent clusters’ to facilitate analysis. This report uses data from LinkedIn’s cluster of tech skills.

LinkedIn member profiles provide skills information directly from the individuals with those skills. LinkedIn data can provide a highly granular picture of skill supply and the relationship that the member who holds these skills has with individual roles, sectors, and functions. This same picture cannot be painted using survey data, owing to small sample sizes. Moreover, LinkedIn data allows members to describe their skills in their own terms – a bottom-up approach. In contrast most surveys take a top-down approach, by asking employers to select skills from a short and pre-determined list of terms.

LinkedIn insights provide a highly granular picture of the skills that individuals have, and the jobs they do. Existing methods of capturing the jobs and sectors that make up the UK economy, like government surveys, fail to capture granular information on skills supply and emerging forms of work. Small sample sizes, a high level skills taxonomy, and the lag in producing insight from the survey – it typically takes a year from field to findings, and the survey is conducted every two years – potentially leads to misplaced prioritisation of policy initiatives, and an outdated understanding of the UK’s skills priorities. In a political climate defined by uncertainty, the challenge of augmenting the information we have on skills becomes ever more crucial.

LinkedIn data is used by employers to make decisions on recruitment. As such, the data is in terms that employers already engage with. As representative as large scale surveys might be, employers are unlikely to use this to drive hiring, or workforce decisions. LinkedIn insights provide a current picture of the labour pool in the UK.

But this data has some limitations. As a result, the findings should be seen as indicative and any percentages reported (for skills, skill-combinations and regional skills dynamics, for example) should be treated as approximations. Some considerations to take when using LinkedIn data include:

• Not all individuals use LinkedIn There are a number of reasons why an individual may chose not to curate a public profile on LinkedIn. Alternative platforms are also used by individuals to engage both socially and professionally, meaning that an individual may present information on the skills that they hold elsewhere, and chose not to use this provider.
• LinkedIn members may choose to use the platform for different purposes, and the information they choose to make public will be influenced by a variety of factors. Information that is entered by LinkedIn members on their profile page is influenced by how the individual chooses to use the site, which can vary based on professional, social, and regional culture. These variances were not accounted for in the analysis.
• The skills that LinkedIn members have on their profile may not accurately represent their skill-set LinkedIn members may exaggerate the skills that they hold on their profile, in order to appear in a wider range of searches, increase their appeal to potential employers, or appear more professionally competitive compared to their peers. On the other hand, members may not update the skills on their profile regularly, thereby understating the skills that they hold. LinkedIn does not verify the qualifications individuals say they have. However, as a public profile of an individual’s academic and professional achievements, on a professional site, (and used by recruiters to identify candidates,) it is likely to be in the interests of members to accurately represent themselves.
• LinkedIn member profiles are not exhaustive A member may not mention every skill that they hold. This is particularly true for basic skills that individuals may assume others will take for granted they possess. Take a software developer with C++ competency, for example, it would seem unlikely that this individual would suggest they are also competent with basic internet use, despite the high likelihood that they will be.

Data points in this report are aggregated to ensure that they do not disclose any information on individual members. At LinkedIn, a core value is ‘members first’ – acting in the best interest of members. As such, while there are significant efforts to accurately express the information contained within member profiles, whilst ensuring that personal, or individually identifying information is not disclosed. Information has been included where data has been excluded, and the caveats associated with this are explained. It should be noted that all information is, as much as possible, an accurate reflection of the LinkedIn network.

We define tech skills as the ability to do something productive using digital technology, and in this report we’re using this definition with reference to work, rather than everyday life. Examples of tech skills range from Java Development, and Application Packaging to Computing – showing the breadth of tech skills. LinkedIn collates thousands of unique terms that members use in their profiles to describe a whole host of attributes like knowledge, abilities, behaviours, work activities, and skills as conventionally understood. Members indicate the skills they hold within the skills section of their online profile. LinkedIn clusters the individual skills that members choose to display into categories for analysis, approximately two hundred skill clusters. Each skill cluster is then further aggregated into ‘parent clusters’ to enable analysis. LinkedIn’s analysis is based on members whose skills fall into the tech skills cluster.