Leading data across more than 30 countries
What running EMEA data, reporting and analytics at HSBC taught me about governing data across more than 30 countries: one governance model, a blended team, shared platforms, and savings that recur.
Quentin Casares · · 3 min read
A regional data role in a global bank looks, from the outside, like a bigger version of a national one. It is not. Across more than 30 countries the hard part is rarely the technology. It is that every country has its own regulator, its own history and its own sensible reasons for doing things differently, and a regional function has to create enough common ground to be useful without flattening what each market genuinely needs.
One governance model, written down
I built the regional governance model on DAMA-DMBOK. The value of an established framework in a multi-country setting is not that it is perfect. It is that it is neutral. No country can claim it is another country's way of doing things imposed on them, and every data owner can read the same definitions of ownership, quality and stewardship.
A common model also changes the conversation with local leaders. Instead of debating whether a country's practice is right, you can ask which part of the shared model it meets and which part it does not, which is a far easier discussion to have and to close.
Shared platforms make the model real
A governance model on paper does not survive contact with 30 countries unless people can see it working. We implemented Collibra for governance and Oracle OBIEE for reporting on a data lake, and led a PeopleSoft rollout supporting 300,000 users. Shared platforms gave the model somewhere to live. Ownership and definitions sat in one place, and reporting drew on one estate rather than on a different extract in every market.
The PeopleSoft rollout is a reminder that data leadership at this scale reaches well beyond analytics. A rollout to 300,000 users only lands if the data underneath it is owned and trusted.
A blended team needs one rhythm
The 40-person team was blended, and I also directed partnerships with Accenture and Deloitte. Leading a team like that is mostly about rhythm. Everyone needs to know what the regional priorities are, how decisions get made and how work moves between in-house staff and partners. Without that, a blended team quietly becomes several teams that happen to share a manager.
I also led GDPR technical compliance for the region, which tested the same muscles. A regulation that applies everywhere but lands differently in each market needs one owner, one interpretation and a clear route for local exceptions.
Savings that recur, not savings that are announced
The £2m+ in efficiencies were recurring, and that word matters. One-off savings from a reorganisation are easy to announce and easy to lose. Recurring savings usually come from removing duplicate effort for good, which is why common governance, shared platforms and one team rhythm matter commercially as well as operationally. They are how a regional function pays for itself year after year.
What I would ask of any multi-country data function
If you run data across many countries, or are about to hire someone who will, ask four things. Would every country's data owners recognise the governance model as theirs? Do the main platforms serve the whole region, or does each market quietly keep its own? Does the blended team, partners included, work to one set of priorities? And can you point to the duplication that was removed to create the savings? A function that answers yes to all four is regional. One that cannot is a federation with a shared org chart.
If you are building or steadying a multi-country data function and want an experienced view, book a confidential introductory call.