<img height="1" width="1" style="display:none;" alt="" src="https://dc.ads.linkedin.com/collect/?pid=306561&amp;fmt=gif">

CDAO New Zealand

Auckland, 4-5 November, 2020

Interview with CDAO NZ speaker:
Kevin Sweeney, Principal Advisor - Data leadership & Capability, Stats NZ

Q1. What is the biggest challenge you face within your role today and how are you looking to tackle it?

My work and that of my team involve fostering change across the New Zealand (government) data system, primarily through publishing guidance and providing consulting support to help agencies lift their data capability. Following the COVID-19 pandemic, the potential for change reflecting a sense of immediacy and urgency, and with a decidedly system-level focus, has become commonplace.

But how to shift agencies, operating under a model that has them functioning in service of siloed agendas for the most part and within what are often lengthy timeframes? As one means of addressing that, we are currently developing a set of recommendations to increase the resilience of the government data system, based on lessons learnt during the pandemic. Equipped with this evidence, of both problems and successes agencies experienced, we hope to be able to make a strong case for potentially significant, and in some cases, rapid change.


Q2. Trust, privacy: What is required to provide reassurance/ trust that government-held data is being managed and used responsibly and safely?

Put simply: transparency. A simple concept, but one that can cover a range of things associated with trust. Transparency can represent the regular publication of publicly accessible work that allows those outside of government a view of what is being done. It can represent the design and implementation of infrastructures like data flow maps or business process models that also provide a view into the way the business of data is conducted within the organisation.

Attitude is also a critical part of delivering transparency, because it represents a way of doing things that requires commitment. In many cases operating transparently is going to contrast with traditional government, which is often fuelled by risk models and characterised by a culture of risk-aversion. As with any shift to openness, transparency incorporates a level of vulnerability and that can be a hard sell. It also implies a highly customer-centric focus, which involves investment to truly understand the customer’s need for trust and assurance, and therefore their motivation for demanding high levels of transparency. But successful adoption of that outside-in perspective is critical if efforts at transparency are going to deliver needed assurance.


Q3. The future of data and analytics post-COVID-19. How have businesses had to change their approach around systems and frameworks?

My team is currently developing a set of recommendations to increase data resilience in the New Zealand government data system, based on lessons learnt from the COVID-19 pandemic. As part of that work, we’ve collected stories, both positive and negative, from a range of agencies. These experiences exhibit common patterns, or themes, that are useful for characterising the sorts of changes that have happened. And per a wider scan of the literature, the themes in the New Zealand government context appear to mirror those experienced by governments around the world. The areas of potential change include:

  • the need for agility
  • a shift to fit-for-purpose data quality
  • the establishment of a central data authority governance
  • data that reflect communities and localised geography
  • availability of trustworthy and highly ingestible statistical information.

These themes represent drivers that have been triggering and will continue to influence the changes that are needed as the COVID-19 pandemic continues to evolve. The over-arching approach is one of resilience, where we position ourselves to increase the likelihood that data and related infrastructure can help us deal with any future circumstances in an efficient and effective manner.

Q4. What does a good data governance structure look like: How are you building sustainable data governance?

In many ways the sustainability of a governance framework is dependent on relevance, so its structure should be designed to facilitate that. By its nature governance, particularly that implemented using traditional top-down and hierarchical approaches can represent a vague concept to those who aren’t directly involved in governance groups. This lack of relevance is often the source of failed governance frameworks, relegating them to the proverbial shelf.

While capturing and delivering the insights of those at senior levels is an important requirement of a good governance framework, likewise is ensuring that those inputs are easily and readily operationalised to influence data practice and the work of all staff with data responsibilities. Data governance establishes or reinforces organisational norms in regards to data, and as such it must permeate all levels of the organisation in a holistic and meaningful way. Therefore, a combined top-down and bottom-up data governance model is needed to be sustainable.

In addition and factoring in the undeniable influence of the COVID-19 pandemic, good data governance is going to contribute to resilience. It must be designed to be agile, easily shifting into a different configuration for instance during the response phase of a crisis. Experience has taught that crisis governance is best centralised with an agreed authoritative source and clear accountabilities for others. The use of a central data authority can save precious decision-making time and ensure everyone working with data is delivering to a common goal, for instance supporting data sharing and data-driven decision-making. Following the response phase, this governance structure should be designed to convert back to a “peacetime” mode, with the ability to support any changes suggested during the initial response.

Q5. What frameworks should be used?

We have developed an operational Data Governance Framework (oDG) that we are promoting across the government data system as a supplement to traditional top-down governance models. When used in conjunction with traditional models, this provides a combined top-down and bottom-up approach to data governance which we call holistic. Holistic in this sense means that data governance is relevant at three levels of an organisation: executive, management and operations. So it supports the manifestation of data norms at the three levels, each of which will have unique needs of data governance, while simultaneously facilitating joined-up data governance that’s positioned across all three levels and supporting the enterprise.

Under a holistic approach, all of the elements facilitated by a good data governance framework – data-informed decision-making, robust data and metadata management, agreed data quality, data security, data sharing agreements, etc – are moving as needed up and down through the executive, management and operations environments to facilitate good data practice.

This works in conjunction with a data as strategic asset and whole of lifecycle view of data, such that any staff member is well-equipped with actionable knowledge of the data asset and clarity on what constitutes good practice, in their specific context and while they have responsibility for those assets. Likewise, it allows them to add to the knowledge of that data asset and contribute their inputs (collection, quality control, analytics) in a way that is enriching, reflects organisational norms and is easily accessed and leveraged by other staff downstream in the data lifecycle.

Watch on-demand