Creating the data-driven public sector organisations of tomorrow

Across the public sector, agencies and departments need to become more data-driven in order to improve efficiencies, combat fraud and improve the citizen experience. At the same time, ensuring data privacy and protection must also be a key priority.

As they continue on the digitisation journey, public sector bodies can learn a lot from the experience of the financial services industry. Given the significant cross-overs that often exist between datasets held by banks and, for example, HMRC or Companies House, there is also significant scope for greater collaboration between them.

The financial services sector has itself been on a journey to become more data-driven. One area of real focus has been KYC processes and anti-money laundering controls, where regulatory requirements have been steadily rising in order to combat financial crime.

At KPMG, we have worked with a range of financial institutions to help them enhance their monitoring and controls processes. Institutions already had significant controls in place, but these tended to be quite manual in nature, often requiring many hours of staff time to conduct a thorough customer review. Once a customer had been checked and onboarded, there may not have been another full review for several years – but this was proving insufficient to effectively manage financial crime risk. Over time, a company’s trading activities, ownership and structures may inevitably change – usually entirely legitimately but sometimes for less honest reasons – and this may not be fully reflected either in a bank’s own information or in public sources such as Companies House.

There was a realisation that the KYC approach needed to become more event-based, whereby data sources and information could be ingested on a more continuous basis to identify changes when they occurred, something that has become known in the industry as ‘Better KYC’.

To do this, we helped several banks create Big Data environments that pull in and monitor data from a wide range of internal and external sources. Organisations often get hung up on data accuracy and this can be a blocker to transformation initiatives getting started. But one of the key principles was that not all of your data needs to be 100% accurate for it to be monitored and changes observed. After all, once an issue has been identified, it can then be delved into more deeply at that stage.

The environments we built involved monitoring and comparing across sources, so that information given by customers can be compared to data available through social media, news reports, credit rating agencies and other databases.

Probably the most significant learning was that, to look at real-world activity, you have to let the data structure itself rather than force it into the format you are imposing (an Excel pivot table for example). To do this, you need to use the right technology. The environments we built utilise Graph technology, which identifies patterns and correlations in the data. You can input what to look for, but it finds the patterns and presents them itself. This can really open your eyes as to how analytics could and should be done.

The results have meant that it now takes significantly less time to conduct customer reviews. Processes are more robust, monitoring is more dynamic, but it’s also taking up less human time. At one bank, there was a 400% uplift in potential financial crime indicators.

These principles and techniques could all be applied to the public sector, across tax, benefits and financial databases, patient information in the NHS, environmental and ESG reporting information, and more. Investment is needed – but the returns are clearly there.

Privacy is a key issue, but can be effectively managed and maintained if the right governance and controls are embedded. Furthermore, it’s an area where we will significant developments over the coming years through new highly secure technologies such as NFTs, biometrics and blockchain. Take a new company formation, for example. The model of the future could be that when a new company is created and registers with Companies House, an NFT is generated. This NFT is used across every aspect of the company – for its VAT registration, its bank accounts, its tax records etc. The secure NFT connects everything together and is the identifier.

In addition to this, new forms of encryption are already being created such that data processors can access information and run analytics over it without actually being able to ‘look inside’ and see the identity. It is possible to share data but retain privacy.

This is the way the future is heading, and it has huge ramifications for public sector and government bodies. They need to keep up their determination to become data-driven organisations, learn from and collaborate with the financial sector where relevant, and keep their eyes fixed on the prize of better KYC that also delivers better experience for citizens.

What are the key considerations when moving to a controlled data focussed environment?