Why Trusted Data is Essential for Data Democratization

The Q&A with Santiago Castro, Link Group

Automation is being driven across the enterprise with greater frequency. But for automation processes to be successful, they must be powered by trusted data. For example, for a financial trading application to be fed real-time updates to monitor a portfolio, the underlying dataset must be accurate. So how do you ensure that trusted data is available to the people who need it? And how can you upskill the staff to make sure they make the most of having access to it?

These are some of the topics raised at the recent UK & Ireland CDO Executive Summit in London.
One company all too familiar with the trend is Link Group, the largest services provider in the superannuation administration industry in Australia. I discussed the power and challenge of automation, as well as the rise of the data marketplace, data citizen, and more, on stage at the Summit with Santiago Castro, Global Head of Data and Analytics, Corporate Markets, at Link Group. The following is an excerpt.

Ajay Vohora: How are you using data to enable more automation within your organisation?

Santiago Castro: Thinking about data-driven automation should be anchored around end-to-end business processes that drive a business scenario and directly improves a business outcome. Even the discovery, classification, and management of data itself can be automated to accelerate time to value achieved from applying data.

In the past, if a business manager wanted access to data or even wanted to understand what data was held within the organisation, they had to speak to someone in central IT, who then had to perform complex manual searches and then start feeding the information back. Now, automation can be applied from the start, removing this manual overhead. Data automation tools can transform siloed and fragmented data into an accessible resource and a single version of truth for people to start working with.

Santiago Castro

Now these business users have trusted data, they can put automation into action themselves. Take operational reporting as an example. Organisations are still pulling together multiple Excel documents, manually copying and pasting things together to custom build reports on a regular basis. With automation, you remove that repetitive process. The data you need gets pulled together and shared with you according to whatever reporting schedule is required. Typically, a business user is relying on reports to identify thresholds that are being reached, or to spot exceptions that need to be addressed. With automation, the business data user can be notified as soon as those thresholds are met – by something as simple as an email arriving in their inbox to inform them.

With trusted data combined with Robotic Process Automation (RPA), you can go on to look at how every manual process – not just reporting – can be automated.

Vohora: How do you define data democratisation?

Castro: For me, I view it as data being available for everyone. If an organisation wants to mature in its data journey and use data to inform decisions, then senior leadership and executives all need to be invested in this goal. You don’t need to be a mathematician, data engineer, or Python programmer to add value – software driven DataOps has evolved so any business data user can plug into systems and play around with their own data. We are already used to applying digital tools, apps, and automation across all areas of the enterprise – and automating the management of the data used to drive the business outcome should be no different.

Ultimately, an organisation should be heading towards a workforce of data citizens, supported by data champions. Everyone should be able to use data, supported by champions who have the know-how to help individuals get the most out of it. Data champions can be critical in sparking a chain reaction amongst staff, driving wider adoption and enthusiasm towards new tools.

The challenge lies in changing the culture accordingly. Teams need to be trained and upskilled in a manner that allows them to use data effectively and confidently.

Vohora: How does trusted data enable you to achieve data democratisation?

Castro: Organisations should strive to deliver a ‘data hub’ or ‘data marketplace’ – a single view of data that presents data wherever it exists and describes data – with metadata – that can be searched, visualized, and navigated. The role of the Chief Data Officer should be in delivering high quality data to this hub for people to access. Once you have trusted data in an accessible fashion, people can begin to self-service their data and insight needs.

This also offers economies of scale that really sell the benefits to internal stakeholders. Business teams only need to invest once in creating this ‘data marketplace,’ but it can be re-used again and again. Any data models that get built within it can again be registered and then replicated across teams to gain the insights they may require in their specific business domain.

Vohora: A data marketplace like this really allows reusability of data products and helps improve the return on investment. The key factor in enabling it lies in the data’s provenance, quality, integrity, and validity. Trusted data is an essential pillar at the core of what we are enabling our customers to achieve.

Data has entropy: It will always proliferate based on how it is generated by system events or human activity and then shared. Older legacy IT systems are often built as a closed loop where data is siloed. Fragmentation of data siloes makes it inaccessible. Leaving data where it is but automating a single version of truth across these data siloes and applying common data definitions with accurate proof of provenance, allows business users to have confidence in the data they need and how it can be applied to achieve a business outcome. Moving all data to a single data store is definitely not required. Instead, generating a single metadata view across silos has a faster return on investment.

Of course, good data hygiene needs to be factored in. The data that drives the business outcome and customer experience needs to be managed cyclically to ensure it remains high quality – and is improved over time.

Automation for all data operations relies on trusted data, presented as a data marketplace. Achieving this gives business data users confidence that any analysis or subsequent decision-making will hold true over time. We see this truly working when data citizens are able to collaborate, to design, and to build a better customer experience from their data – and their imagination.   

Ajay Vohora is Vice President, Digital & Data Products, Hitachi Vantara

Be sure to check out Insights for perspectives on the data-driven world.