As data management practices mature and evolve, I sometimes think that the end goal is being lost.  There has been significant talk in the recent past about data management topics of gold copy, governance, cloud, data mesh and data fabrics, machine learning, AI and data science etc.  All worthwhile.  However, any of these endeavours will be considerably less successful if the underlying data quality is not established and maintained.  After all, garbage in, garbage out.

I was talking on a conference panel recently and was given the question of ‘Where do we start?’ in relation to Enterprise Data Management.  I opined that data quality should be the starting point and then build from there.

Firstly, identify your crucial data sets and elements.  Then measure them for data quality.  This cannot be a one-off exercise.  Ideally, it should be done as regularly as there are updates to those datasets.

As part of the exercise of quality checking and measuring your data, it may also be an appropriate time to add data governance to those sets.  Governance attributes relating to the set, including definitions/dictionary, stewardship, lineage etc., can be added.  If you find that your data quality is problematic, then look to remediate your data sourcing, integration, and mastering logic/processes.  After all, ‘Completeness’ is a data quality measure, and one that may require you to make changes to parts of the Enterprise Data Management process including the ETL and MDM components.  However, if you start with a data quality lens, it should help ensure you target the most important data sets first and identify weaknesses.

When selecting a solution for your data quality management there are several capabilities which we at CDSL felt were important in the development of CuriumDQM to provide a comprehensive DQM solution.  These include:

• Test-in-situ – avoid time, cost and risk of moving data around as a precursor to validating it
• Exception detection – quickly configure and manage data quality checks
• Exception Management Workflow – monitor and manage data exceptions and end user queries from inception through to resolution
• Management Information (MI) Reporting – ensure understanding of both current, and historical DQM activities, including libraries of quality checks, data volumes, frequency and quality scores
• Complete Data Confidence – a dashboard to provide clear understanding of data quality at a specific point in time
• Reduced Costs of Errors – trap errors quickly and remediate them with the help of integrated enquiry tools to underlying and related data.  Automate notifications.  Use quality scores as a staging gate before additional processing occurs
• Amendment Capability – have the option to update/create data manually where required with appropriate permissions and audit trail embedded
• Exception or Data Analysis modes – generate either data quality exceptions, tickets or quality exception profile information in relation to datasets

Data quality management should be a cornerstone to any mature data management capability within an organisation.  If you can’t measure something, then it is difficult to improve.

APRA’s CPG 235 provides an excellent set of best practice guidance and is a ‘must read’ for anyone involved in data management in the finance industry, here in Australia, in particular.  Whilst more broadly concerned with data risk as whole, much of this guideline outlines expectations in data quality frameworks.  Currently it is a guideline and not a standard!

However, if (or when?) APRA regulates enforceable standards over data management, as other regulatory authorities around the world are doing, it is safe to assume that CPG 235 would be one of the bases for that standard and there will be a mad scramble for firms to be compliant.

But why wait?  CuriumDQM is a sophisticated DQM solution that provides a range of capabilities that CDSL has developed for over a decade whilst servicing global asset managers, pension funds and market data providers.  Curium is exceptionally quick to configure to your own datasets, to see results and a return on investment.  When combined with CuriumEDM (Master Data Management and Governance), it provides demonstrable compliance with data management best practice guidelines such as CPG 235.  And that’s just good practice!

If any of this resonates with you and you would like to discuss, please don’t hesitate to reach out to your local CDSL representative.