Research and analysis

Data foundations and AI adoption in the UK private and third sectors: Executive Summary

Published 16 August 2021

About this research

DCMS appointed EY to conduct an evidence analysis and primary market research to assess the extent of data foundations and AI adoption. In addition, our research covered the impact of, and barriers to adopting data foundations. To inform DCMS’s goals of helping build a world-leading digital economy that works for everyone, this study sets out points of view from organisations across the UK economy — including third sector and small and medium-sized enterprises (SMEs) — on the perceived value of data in decision-making, the adoption and use of data foundations and artificial intelligence (AI), barriers to the adoption of data foundations and the key considerations for Government to address these challenges.

Data foundations is defined in the National Data Strategy as data that is:

• Fit for purpose
• Recorded in standardised formats on modern, future-proof systems
• Findable, accessible, interoperable and reusable (FAIR)^{[footnote 1]}

Summary of key findings

The overwhelming response from participants suggests that data is deemed important to the success and growth of organisations across the private and third sector. However, some industries (e.g., Life Sciences, Finance, Industrial Products) expect to derive greater value from improved data foundations than others (e.g., Services and Infrastructure). Government support in helping industries realise greater value from data foundations could positively impact the UK’s gross value-add (GVA).

Common challenges around the adoption of data foundations included the availability of staff with relevant data skills, challenges with legacy infrastructure, and lack of funding. These were identified across all sectors and industries of the UK economy. Data-driven interventions could include encouraging organisations to redeploy funds, with a focus on improving data foundations adoption and supporting new job market entrants and experienced professional retraining for more data-enabled, technically focused roles.

Cultural challenges and obtaining buy-in from management are less common issues, which indicates that organisations broadly accept the need to adopt and improve data foundations to operate their businesses in the future successfully.

Our key findings in the three areas of focus are summarised below.

1. Value of data

— 99% of participating organisations agreed that data is important to their success, with 90% of respondents having a data strategy or data-related initiatives in place.

— Organisations expect to realise value from their data strategy and data-related initiatives mainly through increased productivity (60% of respondents), cost reduction (47% of respondents) and improved customer engagement (46% of respondents).

— There were no material differences in the perceived value of data (as measured by the Perceived Value of Data score) between organisations of different sizes, age, AI Adoption Level or geographical location. This varied from the evidence review, in which we identified that larger and younger organisations were more likely to understand the importance of data.^{[footnote 2]} This suggests that the understanding of the potential importance of data has become more widespread.

— Although organisations understood the importance of data, we found they have challenges quantifying the value of data, return on investment, and the impact of data improvements. This results in organisations being unable to assess the effectiveness of data initiatives or prepare compelling business cases for investment, which may be constraining organisations’ investments in data and data technologies.

— Quality was overwhelmingly identified as the most important data characteristic to an organisation’s success, selected by 41% of survey respondents. This was consistent across organisations of different size, age, sector, AI Adoption Level or in different geographical locations.

— There were no material differences in the most important data characteristic between organisations of different sizes, age or sector. However, significant differences were identified between industries.

2. Adoption of data foundations and AI

— The adoption of data foundations appears to be relatively widespread, with no significant differences in the Data Foundations Adoption score^{[footnote 3]} between the size of the organisation, region or industry. However, the level of data foundations adoption in the third sector was found to be relatively low compared with the private sector, consistent with the findings of our evidence assessment.^{[footnote 4]}

— Organisations are still at a relatively early stage of their data journey, with many organisations focusing their data strategy on improving data quality and governance (63% of respondents), security (53% of respondents), and data sharing and usability (47% of respondents). However, in our opinion, the greatest value will come from using data to inform responses to genuine organisational challenges and opportunities, and that is still some way off for many organisations.

— Organisations see benefits from improved adoption of data foundations and expect a wide-ranging positive impact, including increased productivity (80% of respondents), revenue generation (75% of respondents), and customer engagement (72% of respondents).

— AI remains an emerging technology with 27% of organisations at Released and Advanced level; 38% of organisations planning and piloting the technology; and 33% of organisations neither having adopted AI nor planning to.

— 56% of respondents are planning to increase investments in AI technologies within the next three years, and only 2 out of 399 survey respondents stated they would decrease investments.

— Adoption Level is significantly higher in the private sector, with 70% of private-sector organisations planning or already using AI, which compares with 42% in the third sector.

— Within the UK private sector, 90% of large organisations have planned or already adopted AI, compared with 48% of SMEs.

— From an industry perspective, organisations operating in Finance and Technology, Media and Telecom (TMT) report the highest levels of AI adoption, with 52% of respondents from the Finance industry and 38% from the TMT industry being at the Released level (i.e., AI is put to active use in one or a few processes in the organisation) or Advanced level (i.e., AI is actively contributing to many processes and enabling more advanced tasks).

— There was no linear relationship between adoption of data foundations (as measured by Data Foundations Adoption score) and AI Adoption Level. However, organisations with higher AI Adoption Levels also had a higher Data Foundations Adoption score, indicating that data foundations are a necessary but not sufficient condition for adopting AI.

3. Barriers to adoption

— The key barriers preventing organisations from adopting and improving data foundations are:

• Lack of skilled personnel (14% of respondents identified this as the single biggest barrier)
• Challenges with existing infrastructure (14% of respondents identified this as the single biggest barrier)
• Lack of funding (11% of respondents identified this as the single biggest barrier) $CTA

— Cultural challenges and lack of management sponsorship and engagement were the least common challenges (i.e., the most absent of all barriers selected) reported by organisations.

— Frequency of occurrence of barriers, their impact and how they may evolve varies across industries, suggesting barriers to data foundations adoption are industry dependent.

— Barriers also appear to be dependent on the level of data foundations adoption (as measured by the Data Foundations Adoption score). The key challenges (i.e., most frequently selected) for organisations with a relatively low Data Foundations Adoption score include lack of skilled personnel and management buy-in. Conversely, organisations with a relatively high Data Foundations Adoption score reported challenges with existing technology infrastructure, risk of disruption to the organisation, and data-related regulation.

— 68% of respondents agreed that the Government has a role to play in helping organisations use data more effectively. The key data-related Government initiatives respondents would most welcome were:

• Investing in providing people with data skills and improving access to workforce with relevant data skills (63% of respondents)
• Providing funding to support effective use of data (38% of respondents)
• Investment in improving and releasing datasets (37% respondents)

— Based on interviews, there are many instances where public sector data is already available, but it is of varied quality and often not in an easily accessible, usable and consistent format, making it challenging to use by the private and third sectors.

1. https://www.gov.uk/government/publications/uk-national-data-strategy/national-data-strategy ↩

2. DataKindUK and Data Orchard (2017) Data Evolution Project Report ↩

3. Definition of Data Foundations Adoption score in Methodology section (Section 2 of the full report) ↩

4. Skills Platform, Zoe Amar Digital (2020) Charity Digital Skills Report 2020 ↩