How do you track data architecture standards for better outcomes?

Some of this article seems to suggest a high degree of centeralisation which is not the way in which larger/more complex/data-driven organisations should think about data. It is better to think initially about the data domains and move towards more domain-oriented decenteralisation approaches where domain teams take responsibility for their data with the only "centeralised" elements perhaps being a self-service data platform and governance that is common at an organisational level which mostly relates to notions of interoperability, security, privacy, etc. By doing this, it empowers domain experts to make better use of data but also removes the central data team as a "bottleneck".

Data architecture should be made to fit business needs while still adhering to existing standards such as DAMA-DMBOK or TOGAF. This will ensure not only suitability but also scalability and sustainability, ensuring that whoever take the steering wheel, in a succession manner, can easily understand where to go and how to reach it.

I agree involvement and buy in from all all relevant stakeholders especially those need to comply with is absolutely necessary. But be care reaching out the practitioners too early. I my experience, should speak to them initially to mainly understand the problem space and challenges faced due to absence of any standards then go away and comeback with some good draft to seek their feedback. This is slightly different for sitting with them to draft the docs. This will help to get useful feedback and buyin instead of allowing them to drag you in too much detail.

Stakeholder communication and education is key, especially within organizations that is still unclear on how to measure up the return on investments in Data, including people and infrastructure. Enterprise Data Architecture documentations should be made in several versions (at least 2) varying in level of technical details. Then, share the most generic one to the rest of the organization. This will help especially in providing knowledge on how data is flowing, processed, and stored.

In my experience, delegating compliance to the teams and trusting them that they will comply with is dangerous. Project teams are always face time challenges to deliver their work, and they often try to cut on various things and unfortunately, compliance to such standards is often the victim, so a formal assurance and governance should be done to ensure compliance instead of random audits.

I agree having enablers like processes and technologies will mature data architecture practices. However, one key enabler that precedes both and is usually neglected is People, data-driven environments used to inform our decisions across all walks of life require co-ordination activities across teams to ensure effective and efficient data quality and integrity.