Seven steps businesses can take to maximise data value
We are now in the heart of a digital economy, and some even say we are entering the post-digital era. But how can we be in the post-digital era when most businesses have barely begun to reap the business value of their data?
This is the question posed by Kevin Leahy, Dimension Data's, Data Centres BU Principal Director, who highlights seven steps to take for organisations to get the most out of their data.
Everyone agrees that data is the new oil, but what does that mean for businesses?
"Let's explore this analogy. The price of oil drives global stock markets and the availability of oil drives - not only prices at the fuel pump - but can disrupt industrial production and the movement of goods across the globe," explains Leahy.
He says the oil industry makes extensive use of science and analytics to find new sources and drill efficiently, and carefully manage the supply and quality. But this mean a complex infrastructure to deliver and store across a global supply chain, is needed, and requires protection from natural and unnatural disasters. And all this is happening while we are in the middle of a boom in renewable energy, so oil companies are redefining themselves as energy companies.
"This analogy turns out to be relevant to the story of how all industries are transforming based on data. However, unlike oil, data is exploding and is not supply constrained," adds Leahy.
The three pillars businesses need to build in order to extract the maximum business value from their data are:
* The infrastructure that stores the data, the network that transfers it, and the security/governance to protect it. The complexity of this infrastructure has expanded as a result of the adoption of hybrid cloud and the growth of IOT devices, while network speeds required to access these new services have increased.
* Management of the data. Different from the management of the infrastructure, this is focused on making the data usable across all types of data; including data stored in traditional relational databases and the high growth NoSQL databases, as well as unstructured data. In the past we talked about ILM (information lifecycle management), but when we did, we were largely talking about how to move data to lower-cost storage as soon as possible. Focus was on cost rather than value, and this is changing.
* Getting value from the data. Leveraging insights begins with a view of all the data and extends onwards to analytic projects. While the market was excited about big data a few years back, a wave of pragmatism has taken hold and businesses are now looking for data models to run analytic queries and algorithms against to address a specific business problem.
Leahy says that by addressing only one pillar without the others can be disastrous. And data dispersed throughout a hybrid environment without policy automation could result in exposed data vulnerable to cyber attacks, which could result in loss of data, or create a compliance risk with severe penalties.
"Not optimising the management of the data could result in bad decisions based on old data, and setting a few data scientists off on a journey, could result in massive infrastructure or network spending, or create new compliance risks without proper governance. You need to look at all three in balance."
As a service provider and systems integrator, working with clients around the world on their digital transformation journey, Dimension Data identified these critical success factors:
* Gain visibility of the hybrid data estate, understand the new data available, and the growth rates and gain some immediate insights. While we talk about machine learning and artificial intelligence, the human brain does a great job when presented with data in a visual form.
* Improve accountability and control by analysing and classifying your data. In most cases legacy data has strict policies in place but is missing the classification to link it to other data, for it sits in silos. Classification of that data and establishing the controls for those data class are important steps towards integrating data sources, while staying within governance limits.
* Immediately address increasing security threats, vulnerabilities, and understand the real risk to the business. It sounds obvious but threats are constantly evolving, and many organisations have not changed their processes to keep pace. A simple example of this evolution is that backup copies of data are typically the first target of a cyber attack. Have you air-gapped or hidden your backup?
* Modernise storage infrastructure and exploit cloud storage options to deliver higher speeds and efficiencies. Data value must deliver at the speed of business and most storage infrastructures and networks are not set up to deliver that.
* Address data lifecycle and manage different data types with policies and automation. At the speed data is being created, the ability to keep pace without automation and a platform to ensure the consistent enforcement of policy is approaching zero.
* Minimise downtime of critical business application and protect data wherever it lives. The business impact of any outage is increasing, and backup and disaster recovery in a hybrid world demands new approaches and broader recovery groups. Examine the readiness of your recovery procedures.
* Understand where data is generated and where to place it for maximum value. The old expression is 'data has gravity', and given the volume of data being generated, it is important to understand the decisions that need to be made. If it is a safety decision from an IOT solution in a factory, that will need be processed locally to take an immediate action. If it is a trending analysis that needs to be integrated with multiple data sources, then the data will probably need to be moved to the point of integration. Can the network support get the data to where it needs to be?
According to Leahy when one applies these factors across the pillars of the organisations data strategy, the journey of refining the organisations data in order to deliver business value begins.
"New analytic-based services can present opportunities to differentiate yourself, but without putting the basics in place it will be next to impossible for those services to scale without risk and deliver positive returns through diligent cost management."