By Maurice Titley, Commercial Director for Data and Dashboards at Lumera.
With timeframes measured in decades, regulatory changes and shifting goalposts are inevitable over the lifespan of virtually every pension. Yet even by the standards of our complex, highly regulated and constantly evolving sector, the next five years are set to see unprecedented change across both defined benefit (DB) and defined contribution (DC) schemes.
Digital transformation and the effective stewardship of data will be at the heart of the process. Done well, it’s an opportunity not just to stay compliant, but to future-proof provider systems. Five key areas to focus on are:
Dashboards – I’m connected, so what’s next?
We’re at the end of the beginning for the UK Pensions Dashboard Programme, designed to let savers access their pensions in one place. Around 60 million entitlements are already connected, but two-thirds of schemes are still not onboard as the October 2026 deadline approaches fast.
A surge in connections will happen next year. Once connected, carrying out duties isn’t an exact science. Matching processes will need regular reviews to reflect real usage: for example, when savers don’t provide National Insurance numbers, or typos in scheme data.
To pre-calculate total retirement income across schemes, providers will need accurate, digital data and automated calculations that update annually. Otherwise, they face a near-impossible challenge of calculating this on demand within 10 working days, a model unfeasible at scale.
There’s also the question of whether providers can use match requests to improve their own data. If the verified date of birth from a dashboard search differs from internal records, can the provider confidently update its data?
Getting ready for Small Pots matching
An estimated 13 million sub-£1,000 inactive DC pension pots exist, often without members' awareness. The Government’s proposed Multiple Default Consolidator model will raise significant data matching considerations, as explored in the Digital Systems Feasibility Review, co-authored by Lumera for the DWP.
For example, you might appear as ‘John Smith’ in one scheme and ‘J Smith’ in another. How does your scheme know you’re the same person when transferring your pot?
With an eye on small pots and dashboards, schemes can already work to improve the personal details they hold. The challenge is verifying data for members who, by definition, aren't engaged.
Services using credit reference agency data can help compare records and assign confidence levels. AI models can also assist by identifying likely matches based on past experience.
DB scheme journeys are ever-reliant on data
DB benefits are complex to calculate precisely. As more schemes transfer to insurers, the receiving parties need certainty about future liabilities.
This prompts questions like: How were benefits calculated? Do values reflect all legislative changes? Were records affected by a change of administrator 15 years ago? Are spouse’s benefits current or in need of recalculation?
AI can automate data checks, flag inconsistencies, and help prioritise fixes. Traditional hands-on testing has limitations. AI can spot discrepancies that manual checks may miss, and focus attention where it’s most needed.
Value for Money benchmarking is coming
DWP’s Pensions Roadmap sets out tougher expectations for DC schemes, introducing metrics beyond costs.
A key area is service quality, including data on timeliness, processing speed and accuracy. To be meaningful, benchmarking must reflect a complete, accurate view.
Whether it’s ‘common’ data like personal details, or scheme-specific data, quality must be reported and benchmarked. The Pensions Regulator’s updated guidance encourages assessments across multiple dimensions, including how current and complete the data is.
Making the most of Targeted Support
At present, there's an ‘advice gap’ between people who have access to financial advice and those left to navigate decisions alone. That’s set to change with Targeted Support, which will enable providers to give tailored prompts to members facing key decisions, such as approaching retirement or reviewing contributions.
But success depends on intelligent segmentation. AI and machine learning can play a powerful role, helping identify cohorts based on member behaviour and other data points; analysis that would be difficult to do manually.
The Financial Conduct Authority (FCA) will expect this to be handled carefully. Any analysis done by AI must avoid biases, such as from models trained on old pension products that no longer exist and which shaped customer behaviour in ways that no longer apply.
Gearing up for 2030
Even with so much change around the corner, the industry needs digital transformation that’s pragmatic, accountable and grounded in decades of experience and domain expertise — a Prudent Revolution.
As trustees, pension providers and the rest of the supply chain gear up for 2030, it’s increasingly clear that their data, and how they manage it, will need to transform too.
