This post follows a recent Red Olive round table, at which technology leaders from the insurance sector discussed data governance. Here, we examine the fundamentals of data governance and its implementation within the industry.

Data governance is all about process. It’s about the quality, security, and usability of data. It’s about making sure that data is available – and it’s a strategy for keeping regulators happy.

Perhaps most important of all, though, is that data governance gives us a way to look at the rules, standards and policies that regulate how companies use their data to ensure it remains trustworthy and isn’t used – or misused – in a way that generates risk.

Governance and risk

Ten years ago, what a business did with its data was largely its own concern. That’s no longer true. Today, most industries – but insurance and finance in particular – are subject to regulation and oversight, and data governance has a large part to play in keeping regulators happy.

Data needs to be documented, classified, and catalogued, whether it’s in a data warehouse, an Excel spreadsheet, or somewhere in between. Pretty much every modern application opens or generates data in some way, so it’s essential that any data governance strategy is all-encompassing, covering not only the most obvious data sources, but the business as a whole and every one of its digital assets.

There are internal risks, too, which can be mitigated through effective data governance.
Without it, there’s always a risk of duplicated, wasted effort, which in turn reduces operational effectiveness. Multiple teams may be generating, cataloguing, and manipulating data purely on their own terms, even if it doesn’t match the standards and conventions employed elsewhere in the business.

They might be pushing data into Excel or extracting it from a database for use in a departmental project, without central oversight, and while the fruits of their labours may have some value within their own team, questions remain over the ethics and legal standing of its use. Could the CDO, CTO or systems team – or, indeed, the business as a whole – be sure that it complies with GDPR and other relevant regulations? Without a data governance strategy that considers this point, the answer, at least in the immediate term, will almost certainly be ‘no’.

Governance and revenue

A successful data governance policy doesn’t only minimise risk: it can also help maximise operational efficiencies and facilitate more accurate decision making. It will define how data is gathered, managed, stored, and used, at any point in its lifecycle, giving the organisation, and all employees, a central point of reference against which they can benchmark their use of a source of data. It will provide a means of tracking data flows, monitoring the quality of the data an organisation holds, and deliver a framework through which it can be made available to users across the company.

This democratisation of data, which is key to operational efficiency, is only possible if the business has developed a set of key tools as part of its data governance policy. These may include a data catalogue, reporting catalogue, business glossary, and semantic layer, which, in many cases, will need to be devised and maintained at the development level, in conjunction with the business teams.

Once these are in place, effective data governance is far easier to both implement and sell. Data can drive revenue if it’s managed effectively, since the company can lean on a single version of truth for remunerations, mappings, and other functions that sit at the core of its operations.
Facilitating this is precisely what makes data governance a potential revenue generation tool.

Implementing a data governance strategy

The process of data governance breaks down into many areas, but primarily revolves around the rules, standards, and policies that govern data use.

At a practical level, a good data governance strategy usually enables the following things to happen:

  • Setting data quality thresholds on important data fields, that must be met or exceeded.
  • Monitoring whether these data quality targets are being met and surfacing the evidence with data dashboards.
  • Extending the process to include things like alerting stewards when data fails to meet those thresholds, so that action can be taken.
  • Minimising manual procedures.
  • Providing confidence to senior management that the organisation’s data is being well governed.

Minimising manual procedures is increasingly important since the extent and volume of data within an organisation is continuing to grow, and avoiding manual procedures allows the organisation to achieve a high degree of consistency. At the delivery level, therefore, data governance often relies on automation through the embedding of processes into an organisation’s codebase, catalogue, and APIs.

Yet, while a lot of today’s operations are code-centric, it’s important not to lose sight of the fact that data is more than code: it’s an asset that can be graded, and which can degrade over time. Data governance must therefore also provide some means of considering the quality of the data whenever and however the business encounters it.

Selling data governance

It is not difficult for the CDO or IT department to understand the benefits of effective data governance, but that’s not necessarily the case elsewhere in the business. This is a problem, as without business involvement, data governance won’t work.

How do you get the business to understand what good data governance involves? In many cases, the first step is to outline the approach and make it clear that it’s pragmatic. Start small so you can deliver practical, obvious benefits, then take it from there. Define the rules that data users must follow and, when a big initiative starts, translate those guidelines into data requirements.

Effective data governance isn’t a matter of locking down data and making it inaccessible. Rather, it’s a case of finding a balance. Too much governance risks teams being unable to access the data they need and, as a result, being unable to use it to make effective decisions. Or, they might have to go through multiple layers to sign off on the data’s use, and a modern, agile business simply can’t work like that.

Conversely, a data swamp, devoid of regulation and organisation, would be unnavigable. In many ways, this would be worse. The teams who need to use the data won’t be able to find what they’re looking for, and they’ll be hampered perhaps even more than those who find themselves negotiating with layers of data gatekeepers.

With effective data governance, it’s possible to find – indeed, to define – the middle ground. That’s the place where the needs and obligations of all parties are satisfied and respected, and where the data itself can contribute to the bottom line, without being put at risk.

Effective data governance is key to the safe and profitable use of your company’s data. Whether your business is insurance, or something else entirely, contact us today to find out how you can develop a governance policy that maximises the full potential of the data you hold.