![\"Alex](\"https:\/\/www.dataiq.global\/wp-content\/uploads\/Alex-Sidgreaves-Headshot-200x300.jpg\" "\\"\\"")
\u201cTo date, every data-related problem had received a point-to-point solution, which had worked for many years,\u201d explained Alex. \u201cBut operating in a heavily regulated industry with a complex emerging regulatory horizon, something had to change; particularly as Zurich was diversifying its offerings and growing. There was an urgent need for a data team with the ability to respond to increasingly complex cross functional needs.\u201d<\/p>\n
At this point in time, data provision relied upon specific subject matter experts tied to individual point-to-point solutions \u2013 a model which was complex to operate but had worked well historically. The underlying challenges with the model came to a head when exacerbated by a period of rapid company growth which quickly changed the data needs and meant a new approach was vital.\u00a0<\/p>\n
\u201cThe data team came under increasing pressure with shorter deadlines to deliver,\u201d said Sidgreaves. \u201cAs a business we had gone almost overnight from specific reporting needs from specific systems to trying to pull together huge quantities of data in one format across many disparate platforms and markets \u2013 at scale.<\/p>\n
“Collating everything needed was incredibly complex and time consuming \u2013 simplistically we stress-tested our existing data delivery mechanism and it was not ready to meet these new needs. Data had gone from being very useful to a potential bottleneck.<\/p>\n
\u201cWe started off focusing on risk-based issues, which I think is the best place to start in a heavily regulated financial sector,\u201d continued Sidgreaves. \u201cWe then moved through into a simplification and rationalisation phase \u2013 removing complexity, improving our technology and frameworks and freeing up budget for reinvestment to fuel growth. Now, in the final stage of the transformation, the focus is very much value-based and driving new outcomes that give us competitive edge.\u201d<\/p>\n
Changing the fundamental data operating model and implementing new data strategies takes time, but also requires a lot of active stakeholder management. Historically, the data team had reacted quickly to immediate issues and the organisation was used to being able to lean into this capability.<\/p>\n
\u201cWhen the transformation started, there was a centralised team, but centralised teams are difficult to scale for our business model and aspirations, so this needed to change,\u201d said Sidgreaves.<\/p>\n
The initial centralised data team allowed Zurich\u2019s team to work in a standardised and governed way, providing a single front door for the whole business. Eventually, however, it was noted that the data demands of the organisation had started to outstrip the capacity of the data team.<\/p>\n
\u201cIn effect, the business reached a point where the team is very unlikely to ever be big enough to respond to the whole organisation without some fundamental changes,\u201d said Sidgreaves. \u201cAt the time, we were a team of around 30 individuals supporting up to 5,000 people. You can imagine the amount of data requests that we would get, and not just for one business \u2013 we support seven different brands and multiple business segments, which is a huge amount of demand on one team.\u201d\u00a0<\/p>\n
When this tipping point occurred, the data team found themselves struggling around prioritisation as there was simply not enough capacity to be able to respond to every request. It became the catalyst for change and an overhaul of the way data was used across Zurich.\u00a0\u00a0<\/p>\n
Implementing the change at Zurich\u00a0<\/h2>\n
With the constraints of the existing data team model identified, alongside renewed understanding of the future needs of the organisation, Zurich\u2019s data team was in a position to start introducing new technologies, approaches and systems for providing data-driven insights as and when requested.\u00a0\u00a0<\/p>\n
\u201cUltimately, working in data is like playing chess \u2013 you must be thinking dozens of moves ahead of what you are currently doing to work out where the organisation, the market and the technology is going and make sure the right solutions are there when the business needs them most,\u201d said Sidgreaves.\u00a0<\/p>\n
The team underwent a sizeable transformation, focusing first and foremost on the right skills needed to build a solid future; a process that required a mixture of upskilling and new hires to maintain the right combination of subject matter, expertise and technical depth. The entire data and reporting estate was rebuilt to facilitate easy access to depth and breadth of data \u2013 providing the platform of the future.\u00a0<\/p>\n
This was no small feat and was completed alongside servicing the core day-to-day data needs. It also meant stakeholders across the organisation naturally became more involved in data fundamentals than ever before to support requirements and end user testing. Replacing complex legacy processes meant directly building confidence at all levels that the new approach was not only as good as what came before but offered significant new opportunities. Through the course of the transformation, relationships with the business became far less transactional and more of a partnership.\u00a0<\/p>\n
With a newfound level of data maturity and recognition within the wider organisation, the planned improvements started to come to fruition. There was still a huge appetite for data, but the team and the wider organisation now had the level of cohesion, communication and understanding of their roles and aspirations from data to drive fresh insights. The challenge now was how to continue to change at pace.\u00a0<\/p>\n
\u201cWe had created something which took the company forward, but naturally that success led to wanting to do more,\u201d said Sidgreaves. \u201cFeeding all that pent-up demand through one central pipe was only going to lead to repeating past mistakes. As a heavily regulated industry we rely on a single version of the truth, and there was a risk that this would be lost building a federated mesh system that spread across the business. What we designed was a best of both worlds\u2019 solution. \u00a0<\/p>\n
\u201cThe data team remains a central function with a centre of excellence, but it now runs on capability squads that are positioned within different business functions and running on an agile framework. The business areas are in full control of the what \u2013 they are the best people to do so \u2013 and the data team is in full control of the how as we are the best people to do that.\u201d\u00a0<\/p>\n
\u00a0These capability squads are formed from data professionals who are part of the central data team and fully aligned to architectural best practise and governance. They are placed entirely at the disposal of a single business area who set their change roadmap. This provides a more direct link to data and its potential for the business function, and it allows the data team to be more responsive and understanding of the requests being made.\u00a0\u00a0<\/p>\n
- \n
![\"Dave](\"https:\/\/www.dataiq.global\/wp-content\/uploads\/Dave-Kay-Headshot-200x300.png\" "\\"\\"")
<\/dt>\n- Dave Kay, Lead Data Architect, Zurich.\u00a0<\/dd>\n<\/dl>\n<\/div>\n
\u201cWe have been very clear on one key point: the data skills needed today are very different to what we leveraged a decade ago,\u201d said Dave Kay, Lead Data Architect, Zurich. \u201cThere is a definitive need for deep technical knowledge and diverse skills, and the amount of specialist expertise needed now is time consuming to train ground-up.<\/p>\n
“A big benefit with the new model is that we can deploy these specialist data skills directly into a business function to work in collaboration so there is less delivery latency. By being there with the business functions, we can take a direct steer on what is needed and provide a solid and well governed view on what is feasible. It is not transactional anymore \u2013 it is a partnership, and we are cross pollinating at the same time \u2013 business understanding of data improves, and data understanding of the business comes along with it. We have even seen some success turning actuaries into data professionals and vice versa, which would have been unheard of just a few years ago.\u201d\u00a0<\/p>\n
Mesh and regulation2<\/h2>\n
By creating a halfway point between centralisation and mesh, the data team at Zurich was able to cultivate an approach that was flexible and most importantly ensured a central cross-functional capability to respond to the regulators as and when needed.\u00a0<\/p>\n
\u201cIt was difficult to react to ad hoc regulatory needs in the past, particularly requests on a large scale,\u201d said Kay. \u201cWhat we can do now with one quick query would have taken several people weeks to pull off when I first joined the company. We were very conscious in the new world that the more we pushed towards a fully federated or mesh type model, the more we would go back to the difficulties of the past.\u201d\u00a0<\/p>\n
\u201cRegulatory compliance is needed for our business to operate, but regulatory compliance is not what drives our business forward.\u201d said Sidgreaves. \u201cAs we designed the operating model of the future, the key challenge was to balance the need for strong centralised control, governance and architecture with the need for flexible delivery across our market segments. Too centralised and we constrain the organisation, too federated and we lose the ability to meet regulatory needs. Where the two previously felt like mutually opposed viewpoints, we have managed to find the right balance \u2013 there is a good understanding of what everyone brings to the table and how best to use it.\u201d\u00a0<\/p>\n
By introducing the capability squads around a centre of excellence, the data leadership team were able to develop a level of buy in to data that had not been seen before at Zurich.\u00a0<\/p>\n
The changing face of Zurich data<\/h2>\n
With the evolution of the data team came the evolution of the data strategy. At the start of the transformation programme the data strategy was naturally very internally focused, with a heavy emphasis on data technology. This made sense for the maturity of the business at the time and the need for data tools to power a new data function, but as the model progressed from rapid change to continual delivery the strategy had to evolve.\u00a0<\/p>\n
\u00a0\u201cOur first strategy being technically focused was not a surprise given that we could not meet the needs in front of us,\u201d said Sidgreaves. \u201cBack then we took a risk-based approach \u2013 the risk was there; it was tangible, and it was easy to articulate and understand at all levels.\u201d\u00a0<\/p>\n
As the evolution of the data function continued, stakeholder views began to shift, and the conversation moved away from risk.\u00a0\u00a0<\/p>\n
\u201cRisk was the reason we had to instigate change, but we gradually hit a point where we had done enough that it simply was not the main driver anymore.\u201d explained Sidgreaves. \u201cThe stakeholder view had shifted \u2013 the thinking was that as a team we had put the organisation in a better place and mitigated for the risk in front of us, so we became much more focused on leveraging the solid base we had created for a wider range of value cases.\u201d<\/p>\n
This shift in dynamic meant the team could now directly correlate the actions they were taking to the priorities of the stakeholders, customers and business objectives.\u00a0\u00a0<\/p>\n
\u201cWe are now on our third iteration of our data strategy, and it is very business focused,\u201d said Sidgreaves. \u201cThe new focus is on value generation and how we align to the overall company strategy. There are of course still tech elements to the strategy, as technology is always developing, and we need to make sure we have the right tools \u2013 but the focus has moved. Strategically we are 50% focused on tech, our ecosystem and new differentiators from data science but importantly we are also 50% focused on people \u2013 organisational data literacy.”<\/p>\n
“Data literacy is about empowerment; we want people across the business at all levels to be confident in what they are doing and how best to leverage the opportunities data brings them. It is almost a clich\u00e9, but it really is about building a data culture.\u201d\u00a0<\/p>\n
\u00a0By shifting the focus to data literacy, the Zurich data team hopes that the organisation is better able to talk about data, can have a better interaction with data and lead the way on ethical use of data.\u00a0<\/p>\n
![\"Stuart](\"https:\/\/www.dataiq.global\/wp-content\/uploads\/Stuart-Bevis-Headshot-300x240.jpg\" "\\"\\"")
Stuart Bevis, Visualisation and Insight Manager, Zurich.<\/figcaption><\/figure>\n \u201cThere is an element of having the right strategy at the right time,\u201d explained Stuart Bevis, Visualisation and Insight Manager, Zurich. \u201cThe early strategies were based on building core foundations because they simply did not exist in the organisation. By building the foundations, the organisation can now start to leverage our assets in ways that were previously unimaginable.\u201d<\/p>\n
Change takes time<\/h2>\n
In a fast-changing world where data (and the vendors and technologies surrounding it) are evolving at an incredible pace, it can be easy for a business to expect rapid change from its data function \u2013 but the reality is that change must be a marathon not a sprint.<\/p>\n
\u201cIt is also a maturity journey that the organisation goes on,\u201d explained Sidgreaves. \u201cIf we tried to jump in originally with where we are now, we would have struggled to achieve success. These things take time \u2013 you must build in the right way to be sustainable. Right now, there is a lot happening with generative AI which presents lots of opportunity, but we must be restrained and assess things on merit \u2013 going too fast too soon is one of our key learnings from the past.\u201d<\/p>\n
\u201cData is an iceberg,\u201d explained Kay. \u201cAnyone in the organisation can see the tip of it; dashboards, data science outcomes, insights that improve our business performance. The bulk of what has to be done to make all that happen sits below the waterline, hidden from view \u2013 making data easy for the organisation to engage with requires a lot of expertise and talented individuals to work effectively.\u201d<\/p>\n
As with all data-driven businesses, there is the need to keep talent and provide opportunities to new hires that are more alluring than the competition. This is something that needs to be considered when undertaking long-term transformation projects.<\/p>\n
\u201cAn aspect of our approach now is that we can provide multiple career paths for data professionals,\u201d said Sidgreaves. \u201cIn a fully federated model there is always a risk that talented individuals become stuck, feel there is a glass ceiling or perhaps the team loses them to another business function. We now have the ability to craft proper career paths for individuals that can rotate around different areas of the business and experience everything Zurich has to offer, building a wealth of knowledge along the way.\u201d<\/p>\n
The Zurich data team have worked hard to diversify their talent pool \u2013 from working to foster early careers, apprenticeships and graduate development in the data space to being involved in efforts to encourage return to work after career breaks. <\/span><\/p>\n
A strong belief in the right people for the right role, upskilling and knowledge sharing opportunities has led to a diverse team that is able to draw on a wide variety of highly informed viewpoints. By enabling different avenues to enter the data industry and providing education and skills to non-data professionals, the Zurich data team has also futureproofed the talent pipeline for the foreseeable future.<\/span><\/p>\n
Data is a journey that constantly evolves and reacts to changing business environments \u2013 it is essential that data leaders implement structures and programmes that promote and enable flexibility, but that can be understood by all facets of an organisation.\u00a0<\/p>\n","protected":false},"excerpt":{"rendered":"
Zurich Insurance data leaders discuss their data transformation journey, providing unique insights for developing data operations in a highly regulated legacy business.<\/p>\n","protected":false},"author":19,"featured_media":7126,"menu_order":0,"comment_status":"open","ping_status":"closed","template":"","format":"standard","meta":{"_acf_changed":false,"content-type":"","_searchwp_excluded":"","footnotes":""},"categories":[129,282],"tags":[240,309,308,228,171,237,119,307,179,123,305,310,301,291,85,306],"pillar":[195,193,194],"class_list":["post-7125","article","type-article","status-publish","format-standard","has-post-thumbnail","hentry","category-editorial","category-member-case-studies","tag-case-study","tag-centralisation","tag-data-centralisation","tag-data-culture","tag-data-governance","tag-data-leaders","tag-data-literacy","tag-data-mesh","tag-data-regulation","tag-data-strategy","tag-legacy-business","tag-mesh","tag-regulation","tag-storytelling","tag-strategy","tag-talent","pillar-literacy","pillar-strategy","pillar-leadership"],"acf":[],"publishpress_future_action":{"enabled":false,"date":"2026-06-16 20:55:30","action":"change-status","newStatus":"draft","terms":[],"taxonomy":"category","extraData":[]},"publishpress_future_workflow_manual_trigger":{"enabledWorkflows":[]},"_links":{"self":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/article\/7125","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/article"}],"about":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/types\/article"}],"author":[{"embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/users\/19"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/comments?post=7125"}],"version-history":[{"count":0,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/article\/7125\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/media\/7126"}],"wp:attachment":[{"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/media?parent=7125"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/categories?post=7125"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/tags?post=7125"},{"taxonomy":"pillar","embeddable":true,"href":"https:\/\/www.dataiq.global\/devstage\/wp-json\/wp\/v2\/pillar?post=7125"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}