Data governance - doubts about data

This question around trusting the data goes back a long way in history and yet it’s as pertinent today as it was in 1903 when the Wright Brothers made the first controlled, sustained flight of a powered, heavier-than-air aircraft. Having conducted several test flights prior to their first official flight, the brothers needed to store and interpret that information.

It is remarkable to think of these two brothers being able to identify an issue with the existing coefficients and question the data it was producing. Subsequently, I think we can draw parallels to today.

Numerous organisations across sectors are relying on data as they turn to digital solutions to improve efficiency and their place in the market.

However, CEOs are often sceptical of the data they are receiving in their monthly reports. They are scrutinising the numbers more than ever before, not knowing a lot about the data source and what process led to the final numbers in front of them. The data is extracted from a range of systems and they are put into Excel. With this comes two problems – the ‘save’ and ‘save as’ buttons as you can end up with different sources of truth.

So how do companies create a business procedure to ensure their data is sacred, and with that, trusted? There are several principles and all of them take investment and resources to get right. 

Too often I see a number of people in an organisation access data at the source and it’s uncertain if they have manipulated the data. Knowing that there is an element of uncertainty puts a cloud over the data’s integrity. For there to be rigour around data and establishing data lineage, someone needs to own data in an organisation. Even if a company doesn’t have the capacity to employ a Chief Data Officer, a data owner or data steward needs to be nominated to be responsible for the process in which the data is sourced, interpreted and reported.

If one person retrieves data about an asset and some changes are made, it’s the responsibility of the owner to know where it’s going, what report it’s being added to and how it may join other data to tell or shape a particular story.

It’s also the expectation from management that the people reviewing the data are aware of its source, lineage, how it has been changed and the governance framework set around it.

If you point to a number on a report and ask “where did that number come from?” and no one can tell you, there lies the problem. All data should be able to be traced back to its true origin so that when these numbers are audited or there is a question mark hanging over the data, there is a comprehensive explanation. If this traceability is lacking, the value of the data has been lost.

Not taking these crucial measures has significant repercussions and it can be costly for any organisation. Around the world, there are numerous examples of where bad data has been costly on an epic scale: from elections to financial data and marketing to asset related data. In terms of infrastructure, there have been roads and bridges designed and sometimes built in the wrong location because the data had been changed without being validated against other sources. Mistakes are often made because the data has been tampered with along the way without the end users knowledge.

The key to solid data governance is ensuring a consistent approach to data and the reporting of it. A data glossary should be produced and shared across the organisation so everyone is on the same page. When somebody mentions ‘revenue’, everyone needs to be referencing the same figures and a consistent approach should be implemented to get this result.

In doing so, everyone in the team will be following the same life cycle patterns, applying the same security measures, conducting the same level of auditing, and so too, there will be consistency around retention and recovery of data. 

Establishing data standards is imperative to ensure data quality and its accuracy is maintained. This relates to how reporting is carried out, how data is transferred, captured and published. An essential part of this is determining what data should be kept or thrown away. Some companies keep every data set imaginable but I would challenge this. If data isn’t going to serve a purpose, then there is no need to hold on to it.

My team uses the term ‘data trash’. I know of companies who end up keeping gigabytes and gigabytes of log files that are serving no purpose and they are wondering why their system is running slow. You end up with ‘data trash’ where you've got data and you're keeping it for no reason and mostly likely you'll never use it. If there isn’t adequate cleansing of data, the result can be costly as disk drives and cloud capacity are quickly exhausted.

Our Data & Analytics team does a lot of work with clients specific to environmental outcomes and the rigour around data is essential to meet regulatory standards. Capturing data and the process that has been implemented must be consistent and someone needs to drive a collaborative approach internally to meet the changes in these government regulations. If you don’t, the penalties can be severe and costly for any organisation.

Our Data & Analytics team does a lot of work with clients specific to environmental outcomes and the rigour around data is essential to meet regulatory standards. Capturing data and the process that has been implemented must be consistent and someone needs to drive a collaborative approach internally to meet the changes in these government regulations. If you don’t, the penalties can be severe and costly for any organisation.

Implementing a solid framework to sure-up data governance has the potential to add real value to an organisation. The outcomes depend on how vigilantly the framework is developed and adhered to. Developing a sound governance structure not only provides the whole organisation with assurance around their data, it also allows them to be more agile and to make responsible decisions quickly based on the quality of the data. With this comes efficiency and this is where the benefits can be seen.

A streamlined approach and a trusted one at that, produces savings across a number of cost centres, including corporate assets relating to data storage and retrieval.

Undeniably there is significant work and investment to get to this point but once it’s made and it becomes imbedded in the culture of an organisation, the benefits will not only be achieved in the present, but well into the future.

Developing a road map so you can reach the desired outcomes within a specific timeframe is imperative to a data governance strategy. I’m sure once the results are evident, company decision makers will have no regrets in investing in their data processes.