Context
The Phoenix Group is the UK’s largest long-term savings and retirement business that has continued to grow in recent years through a series of acquisitions and mergers that have expanded the group offerings. With £270 billion of assets under administration, Phoenix Group supports 12 million customers with their pensions, savings and life insurance products to provide the highest quality levels of retirement possible.
Due to the nature of Phoenix’s business, there is heavy regulation, meaning compliance is of utmost importance as the security of customer assets and protection of sensitive data is non-negotiable. Any development to the data processes of the organisation needs to have regulation and compliance inbuilt as a cornerstone while being able to expand and adapt to any upcoming regulatory changes.
In 2021 Phoenix recruited Diane Berry initially as chief data scientist and then transitioned to chief data and analytics officer in 2022. The role involves leading its important data transformation programme to become more data-driven, whilst maintaining the highest levels of data governance. Diane was listed in the DataIQ 100 in 2023, which is the only fully curated power list of the most influential data and analytics practitioners. Since 2014, DataIQ has been tracking the rise of chief data officers, chief analytics officers, data scientists, data governance experts and the leaders of key vendors and service providers. Inclusion in the DataIQ 100 is a notable badge of honour that is widely referenced by the individuals who make the cut.
“Phoenix has an incredible track record around acquisition,” said Diane Berry, chief data and analytics officer at Phoenix Group. “When I joined – from a non-pensions and savings background – the business had recently acquired the Standard Life brand and a heritage company, ReAssure, and the teams within the business were transforming their open growth agendas organically. This is by no means the end of the acquisition cycle as we are an ambitious business, which is where my role comes into play. We have a large number of customers stretched across different entities, and we have different books and different policy admin systems which adds to the complexity of integration and uniformity for the data team as we absorb these new business entities. Data is of course critical to this journey and foundational, so the task became to lead the charge on data strategy for our customers across the whole group.”
Assessing the groundwork
Prior to the development of the new data approach, Phoenix Group was faced with multiple siloed functions, which led to the data team aiming to deliver a single cohesive environment to leverage customer data. The ambition was to leverage the data to improve engagement with customers and deliver the outcomes that customers wanted to improve their quality of life given the nature of the products at hand.
There is an ever-changing business environment and Phoenix Group wanted to stay ahead of the curve and provide quality for their customers. With a customer base stretching from those starting careers to those of retirement age, the team wanted to ensure that we were able to meet the needs of customers, and new audiences and revenue streams primarily within the Standard Life brand.
“The aim was to identify customers moving from one life stage into another,” said Diane. “For example, those in their 40s moving into their 50s and starting to prepare for their retirement options.”
A long-standing issue with a business focused on long-term savings and pensions is that it is very easy to relegate issues to the future problem space, but this is an unsustainable approach that Phoenix Group wanted to avoid by tackling the issue head on. The team wanted to keep up with an ever-evolving customer environment and widen the customer base to include a pre-retirement stage but needed to adapt its data operations and structure to successfully and succinctly make this change a success.
Implementing change
To bring about the change required, the first step required was to source the desired talent and bring them up to speed with the operations of the business and the desired targets.
“From a data perspective, what that meant for the work we were embarking on was to bring in data talent across the firm and bring them all together,” said Diane. “We have done this by bringing talent that already exists across the firm. Then, once the existing talent has been identified, you can bring in external talent to fill in the remaining gaps and build a team for future expansion. After this, as a data leader, you can begin more of a federated approach too which is ideally suited for operations involving multiple brands and teams, but with a common goal.”
Diane continued by explaining that the data team was taking an “eat my cake and have it too” approach by bringing together analytics, data engineering, data architecture, data governance and data talent together to build out into areas lacking capability or dependent on contractors.
The structure
“We needed to build out our data science capability as well as building and establishing our technical talent too,” said Diane.
The way this was organised was through one centralised team, directed by Diane, which collaborated flexibly with all different brands within the group. There were areas of commonality – customer experience, customer engagement, customer behaviours – which allowed a centralised approach to better disseminate the learnings from different brands.
This approach has also enabled the data team to continue working closely with all aspects of the business during ongoing large transformation projects while maintaining its day-to-day operations that align with the broader strategy.
“In addition to the new recruits and team dynamics, we are building a modern cloud data platform in Microsoft Azure,” said Diane. “We work very closely with the broader business to bring different people along for the journey, so they are able to support customers and enhance the organisational data culture.”
Constraints
One of the major concerns for implementing the desired change was available capabilities. There was a worry that the talent was spread unevenly across the different brands that had been acquired and that there was no central team to effectively organise the data projects of the wider organisation.
“Rather than being scattered across the business, the data team needed to be brought together as a collective,” said Diane. “A problem with the existing approach was that data people did not communicate with other data people despite wanting to – they liked working closely with the business, but also wanted to learn from the experiences and perspectives of other data professionals.”
To bring these two worlds together, an evolution of the cloud data environment was undertaken to allow for the pooling of all data across the organisation, but in an equal manner. “We wanted to go through the process of delivery and leverage our data – whether that was as a data lake, a data lake house or a data mesh federated model,” said Diane. “Equally, we aimed to create an environment that really leverages the best from cloud solutions, such as the AI data science capabilities, but also to actively begin conversations around newer ideas such as generative AI.”
It is true that any business will face blockers when implementing a new central system, particularly if that involves migrating from one type of operation to another. What the Phoenix Group team did was take a step back and deeply analyse the wider organisation and the brands within it. “It was about understanding the organisation and understanding what that change management framework in governance looks like,” said Diane. “We run on a specific cycle, so it is about when you need to start making the business case, submitting that business case and going through that designed process to start seeing results.”
This approach has allowed the data team at Phoenix Group to get the initial design of a project in place for the building to commence while providing a great deal of flexibility to non-data professionals that would be needed to ensure success.
“When it comes to change management, we have got a change framework – like many other organisations – that must be adhered to and is underpinned by strong business case,” explained Diane.
With an ambitious change involving technology, tools, daily operating procedures and recruiting more talent, prioritisation within a project of this magnitude was pivotal to success. To be able to manage this effectively, the expertise of a senior leader within the organisation was key. With the guidance of Jackie Noakes, group COO, the Phoenix Group has been able to make significant advancements in the implementation of this project and make good progress in bringing the strategy to life.
“Jackie’s experience with large-scale projects and dispersed teams was invaluable,” said Diane. “For me, it has been transformational having someone come in with experience and say, ‘we 100 percent need to have delivery mindset that is not just about the strategy’. It was an incredible learning experience for me to receive this level of expertise when it comes to delivery.”
One of the major takeaways from this portion of the project was the demonstration that it is not just about convincing the business decision makers to provide the funding but the secondary stage of how it will be delivered to provide swift results and be presented back to the decision makers.
Additionally, the data team needed to emphasise to the business leaders that there was not a precise finishing line for this project. It would continue to evolve and would require ongoing upskilling and expansion as the company grew. This is one of the continuous struggles of data projects as they often do not have a definitive metric to be measured against which can make them difficult to understand for non-data professionals.
Successes
As the project got underway, an immediate win was a shift in the approach due to new talent being continuously added to the team. “It changed the conversation,” said Diane. “When a business colleague is at ideation stage – and there is capability within the team for trying something new – it allows exploration of these avenues. We found we were able to build out our customer segmentation very rapidly and could focus on our priority segments.”
Diane described the team operating in two halves: an offence and a defence. This allowed talent within the team to focus on different aspects of the burgeoning data journey with clear-cut metrics and objectives. The defensive part of the operation would drill into leveraging the data to understand customer behaviours, whereas the offensive part would identify new opportunities and product offerings to set the organisation ahead of the competition. This would ensure that customer needs were not only catered for, but comprehensively understood, and that new pilot projects could be launched to further enhance the propositions being provided and heightening the efficiency of internal data consumption.
Furthermore, in a highly regulated industry, improved data operations will lead to the streamlining of operations, new propositions and heightened compliance.
“For our retail intermediary re-entry, we ran a pilot to prioritise conversations with advisors as supported by propensity models using machine learning (ML) methodologies,” said Diane. “In turn, this supported the decision to launch a scaled operation and we were able to immediately retain an uplift of three times from the previous period. Furthermore, we still have a significant number of intermediary conversations still ongoing.
“In designing our retail strategy, we developed propensity ML models to understand which customers would likely benefit from retirement solutions. Tracking progress to date has shown we are four times more effective in identifying these customers.
“Underpinning our business growth strategy to highlight where we have opportunities in our existing relationships with customers and advisors, we have developed advanced ML segmentations from multiple data sources to guide developments in proposition, marketing and engagements. For example, using the segmentation as a launchpad for a consolidation campaign led to a significant increase of inflows.”
The actions of the data team have seen an increase in the number of active online customers which is improving engagement and digital adoption. This was critical during the Covid-19 pandemic and has encouraged more customers to be aware of how their money is working by having easy-access to multiple online services, including both educational content and transactional services. This has further fuelled brand awareness and consequent surveys have shown an improvement in customer satisfaction and engagement. The rapid successes of these implementations are being spread across the multiple brands operated by Phoenix Group with hopes of continued expansion.
Ongoing monitoring and proactive analysis assist Phoenix Group customers engage with their financial future. This has been achieved by:
- Swift identification of a decrease in product referrals by customers contacted via Telephony colleagues. After targeted refresher training courses, the organisation saw an immediate 50% uplift in referrals, a similar uplift in applications received and an increase of flows retained.
- Monitoring online journeys and identifying opportunities to improve and optimise. This led to an increase of 35% in applications for selected products.
- A series of outbound engagement and test and learn initiatives increased customer digital engagement to 44% for the Standard Life pension customers.
- Customer cohorts that are typically underserved have been identified through use of data science models utilising data from a variety of internal and external sources. This data is being used to inform and improve the financial wellbeing of present and future customers.
A key area of consideration for Phoenix Group was to improve identification and engagement with vulnerable customers. The data findings enabled the development of a first group-wide customer dashboard that gave users insight into customer vulnerability. This dashboard is continuously evolving and following the trends of vulnerability in an ever-changing economic environment.
“This dashboard allows us to determine where we need to prioritise improvements to our existing servicing support,” said Diane.
There is a hard deadline in July 2024 regarding Consumer Duty for closed book products and services and this is one of the near-future projects that the Phoenix Group data team is working on to ensure compliance and quality of ongoing evidencing. There is also an opportunity for the team to improve engagement and product offerings with customers.
“If we look at the next 18 months, there is still plenty to do,” said Diane. “There are a couple of key programmes currently in development and some pilots that are part-way through their trials that have made terrific progress. There is going to be a lot to launch next year as we continue growing organically. We still have a lot to do, but it is a very exciting time to think about future projects and the acceleration of artificial intelligence (AI).”
Future successes
The data team at Phoenix Group is examining its landscape, which has been made more complex with recent acquisitions and legacy operations spanning multiple brands. The end-goal is to be able to implement AI and ML tools to further enhance the customer offerings and internal data operations of the business.
“We need to think about how things such as data mesh could fit in with our landscape,” said Diane. “For example, we have an asset management portion of the business and we will want to mesh some of those functionalities. There are plenty of opportunities right now regarding different forms of technologies and we want to accelerate our AI journey.”
The positive results of the data team’s efforts have led to the building out of data governance across the organisation which has led to increased data literacy requirements for colleagues across the group. This is being monitored via the recent DataIQ Data Literacy survey. “When it comes to data literacy, we have got to do this,” said Diane. “This is going to be a huge resource and we want to shape what we do with this space over the next 12 months. We have expanded our data ownership and have begun embedding data champions and data stewards within the business to develop our data culture.”
Phoenix Group has also established a data council with a dedicated forum which will continue to examine data priorities, quality and future endeavours such as AI. The council will also be advising the organisation on the implementation of new architectures and tools such as data mesh.
“Ultimately, we are always aiming to improve our internal efficiencies, and this is something that is true for all of our departments,” said Diane. “Our data platform and IT colleagues are working closely with experts at Avanade who came in to provide support and expertise on building out these areas and shaping them for long-term futureproofed success. Our platforms are powered by Databricks which has been essential in our processes for improvement across the data space, such as developing our dashboards programmes, and the plans we have for the future due to its ease of use and compatibility. It provides flexibility, a low barrier to entry, scalability and future proofing.”
There is plenty in the pipeline for the data team at Phoenix since implementing its talent change which impacts individual teams, specific brands and the wider group. The work being carried out at Phoenix Group over next 12 months will continue to evolve the way in which data is utilised and harnessed for internal operations and customer experiences, further changing the way that financial products and services are developed and distributed.
Phoenix Group is part of the DataIQ membership programme – the trusted global collaboration and intelligence platform for data leaders. Gain access to the connections, expertise and knowledge required to make better decisions. Find out more: www.dataiq.global/membership/
