Data management strategy: it's time to grow up

Acknowledging the role of data as a key source of economic and social progress, the Australian Government’s Data Strategy sets out the goal to become a “modern data-driven society” by 2030. This includes the plan to establish a national data ecosystem as a key driver of the future economy.

The follow-on question is to what extent organisations will be able to benefit from such measures. In fact, research from Hewlett Packard Enterprise (HPE) suggests that a lack of data capabilities within Australian private and public sector organisations is currently hindering them from extracting value from data — to achieve outcomes such as growing sales or advancing environmental sustainability.

For example, among Australian respondents, only 13% say that their organisation’s data strategy is a key part of their corporate strategy. And only 23% confirmed they have a strategic focus on providing data-driven products or services. Australian respondents also confirmed that almost half of their organisations do not use methodologies like machine learning or deep learning.

The survey is based on a five-level maturity model developed by HPE that assesses an organisation’s ability to create value from data based on strategic, organisational and technological criteria. At the lowest maturity level, called “data anarchy”, organisations’ data pools are isolated from one another, and are not systematically analysed to create insights or outcomes. While at the highest level, called “data economics”, organisations strategically leverage data to drive outcomes, based on unified access to both internal and external data sources analysed with AI and advanced analytics.

The average Australian organisation’s data maturity level is 2.5 — just one step above the “anarchy” level — and only 4% of organisations reach the highest level of data economics. This means that not only are most organisations far from fully leveraging their data as a strategic asset, but they also lack basic capabilities to do so.

Data maturity, therefore, must become a key priority for organisations in Australia, not only because it will be key to advancing the future viability of individual organisations, but the viability of Australia’s overall society and economy as well. There are three key essential principles which must be embraced in order to do so.

1. Swap cloud first for data first

There are no shortcuts towards achieving higher data maturity levels. Doing so requires a holistic change which touches virtually all facets of an organisation. In years past, “cloud first” has been the mantra of CIOs when describing their digital transformation strategy. However, many of them are now starting to understand that they were confusing the means with the ends. A platform decision must always be a function of the outcomes an organisation wants to achieve — not the other way round.

The HPE survey confirms this mind shift. More than half of Australian organisations are concerned that data monopolies have too much control over their capability to create value from data (60%), and 47% are re-evaluating their cloud strategy due to increasing cloud costs (54%), concerns over data security (36%), the need for a more flexible data architecture (45%) and the lack of control over their data (49%).

To holistically advance organisations’ data maturity, they must replace cloud first with “data first” as their North Star of digital transformation — meaning that they align their strategic, organisational, technological and platform choices with the overarching goal of leveraging data as a strategic asset.

2. Embrace an edge-to-cloud architecture

One of the key consequences of this approach is a new platform strategy. Creating value from data requires aggregating data insights from different applications, locations or external data sources. Gartner analysts estimate that by 2025, more than 50% of enterprise-critical data will be created and processed outside the data centre or cloud.

It’s a key characteristic of a low data maturity level when data is isolated within individual applications or locations without an overarching data and analytics architecture. In fact, according to the HPE survey, only 16% of Australian organisations have implemented a central data hub or fabric that provides unified access to real-time data across their organisation, and another 8% say this data hub also includes external data sources.

This is why organisations need to establish an edge-to-cloud architecture that provides the freedom to choose the correct location for their data and applications, while providing one unified model to orchestrate across edges, data centres and clouds. This enables organisations to control their data assets while industrialising their data supply chain at the same time.

3. Know thyself

The first step on a data-first transformation journey should be a relentless inventory — not only of an organisation’s overall data maturity level, but of its specific weaknesses and strengths vis-à-vis its business strategy. For example, a company can be relatively advanced in the use of analytics and AI technology, but its business units are reluctant to share data with each other. In such a case, the greatest need for action is in strategy and organisation. Pilot or lighthouse projects can then be a means of bringing about acceptance and support.

An assessment based on a data maturity model is a solid starting point. This allows organisations to tailor their plans and management-of-change approach to their individual gaps, goals and requirements. Every journey will be unique, and it will be long and difficult. But the reward, if successful, is sustainable profitable growth and sovereignty over one’s own data-driven business model.

Dr. Eng Lim Goh is Senior Vice President, Data & AI, at Hewlett Packard Enterprise (HPE). As principal investigator of the experiment aboard the International Space Station to operate autonomous supercomputers on long duration space travel, he was awarded NASA’s Exceptional Technology Achievement Medal. His other work includes co-inventing blockchain-based swarm learning applications for finance and health care, which was featured on the cover of Nature; overseeing deployment of AI to Formula 1 racing; industrial application of technologies behind a champion poker bot; co-designing the systems architecture for simulating a biologically detailed mammalian brain; predicting predisposition to vaccine side effects by machine learning of gene expression data; and co-inventing a data-intensive fabric for Exascale systems. Extracting value from data is one common factor of all the above. He has eleven US patents, of which four are ESG related, plus two others pending.

Image credit: iStock.com/nespix