Why healthcare transformation stalls
In healthcare, “where to start” is not a planning nicety. It is the decision that most often determines whether a data transformation succeeds or stalls. The instinct is to start with technology, because technology is procurable and visible and feels like progress. But healthcare data carries weight that ordinary business data does not. It is sensitive, it is regulated, and the trust attached to it is hard won and easily lost. Start in the wrong place and the programme gathers resistance faster than results.
Healthcare transformations rarely stall for lack of ambition or budget. They stall because they begin as technology projects in an environment where the binding constraints are governance, consent, and clinical trust. A platform arrives before anyone has settled who may see what, on what basis, and to what end, and the programme then spends its credibility relitigating those questions after the money is committed, when they should have been answered before a contract was signed.
Where to actually start
Start with governance and consent. In healthcare, data governance is not a workstream running alongside the transformation. It is the foundation the transformation stands on. A useful first move is almost boringly procedural. Before any platform conversation, map what data exists, who is lawfully permitted to see it, and the purpose it was originally collected for. It routinely surfaces the real constraint: that a good portion of what the programme assumed it could use was gathered for a different purpose and cannot simply be repurposed. Finding that out in week one is cheap. Finding it out after the build is not.
Start with the question, not the platform. A transformation justified by a tool is a solution looking for a problem. Begin instead with a clinically or operationally meaningful question worth answering, and let it determine what data, governance, and capability the programme actually needs. The platform is a consequence of the question, never a substitute for it. The same discipline keeps business intelligence from collapsing into decoration.
Start where trust is highest. Early wins in healthcare should be chosen for their effect on confidence, not only their technical interest. A first deliverable that visibly respects the sensitivity of the data and produces a real, modest result earns the permission needed for the harder work later. Trust is the currency here, and it is earned and spent in that order.
The POPIA reality
In the South African context, the Protection of Personal Information Act makes the governance-first argument concrete rather than aspirational. We treat the POPIA obligations, lawful basis, purpose limitation, and security, as constraints the design has to satisfy from the first architecture decision, the same way you would treat a budget or a performance constraint. In practice that means consent and access rules are modelled into the data layer rather than bolted on as a policy document nobody reads. That is the job of continuous control monitoring as we practise it: keeping verification and access rules live in the system rather than promised on paper, alongside the security and compliance-by-design that sits in our DAaaS foundation. This is also why we run healthcare work through an embedded model rather than an advisory one, the same instinct behind designing for the realities you actually operate in: governance that has to hold up under POPIA is governance someone has to own inside the organisation, day to day, not a clause in a final report. Compliance that lives in the system is enforced; compliance that lives in a PDF is hoped for.
Key takeaways
- “Where to start” is the decision that most determines success in healthcare data transformation.
- It stalls when it begins as a technology project. The binding constraints are governance, consent, and clinical trust.
- Start with governance, with the meaningful clinical or operational question, and where trust is highest.
- Under POPIA, compliance designed in from the start reinforces the work. Bolted on at the end, it fights it.