I spent three years building technology capability across a group with four operating companies: Bermuda, the Cayman Islands, Guyana and the USA. Before that I worked inside a Caribbean subsidiary of a much larger multinational. So I have seen the multi-opco problem from both ends: from the market that wants to be left alone, and from the group function trying to pull four markets in one direction.
The pattern is the same everywhere. Each operating company has its own currency, its own regulator, its own customer base and its own view of what good technology looks like. Left alone, each one builds a stack that fits its market beautifully and fits the group not at all.
Why the drift is rational
It is tempting to read the divergence as dysfunction. It is not. A market CTO who picks the platform that suits local conditions is doing their job. The Guyana business runs on different economics from the Bermuda business; a vendor that makes sense in one can be unaffordable or unsupportable in the other. Every local decision is defensible. The group-level mess is what those defensible decisions add up to.
Why mandates from the centre fail
The instinctive group response is standardisation by decree: pick one stack, tell everyone to migrate. I have watched versions of this fail more than once. The mandate lands as an unfunded tax on markets that were doing fine, the strongest opco negotiates an exception, and within a year the standard is a document nobody follows. Authority on an org chart does not move infrastructure.
What actually worked
When I had to build a group-wide security capability across all four of our markets, the markets wanted materially different amounts of it. One wanted everything, another wanted two services. The answer that held was architectural rather than political: build the capability once, as one design, and let each market switch on only what it needs. Nobody was forced to carry what they would not use, and nobody had to rebuild later to get more.
That flips the incentive. Markets stop resisting the group standard because the group standard becomes the cheapest and fastest way to get what they already wanted. Adoption stops being a compliance exercise and becomes self-interest.
If you are the one holding the group view
Three things earned more alignment than any mandate. A design authority where the markets' own architects review and shape the shared standards, because people defend what they helped design. Joint sourcing, because a four-market contract gets pricing no single opco can, and money is the one argument every market CEO hears. And a rule that the shared design must serve the smallest market's constraints as much as the largest market's ambitions; the version that works in Georgetown has to be the same version that works in Hamilton.
Unified does not mean identical. The goal is one architecture with local switches rather than four copies of the same stack. Get that distinction right and the group stops arguing about technology and starts arguing about what to build next, which is a much better argument to have.