In the previous Industry Insight in this two-part series, I looked at why we need data governance. In this Industry Insight, I look at how to ensure data governance is correctly and consistently adhered to through the identification and application of best practices.
Wikipedia defines best practice as: "An idea that asserts that there is a technique, method, process, activity, incentive or reward that is more effective at delivering a particular outcome than any other technique, method, process, etc. The idea is that with proper processes, checks, and testing, a desired outcome can be delivered with fewer problems and unforeseen complications. Best practices can also be defined as the most efficient (least amount of effort) and effective (best results) way of accomplishing a task, based on repeatable procedures that have proven themselves over time for large numbers of people."
Given this definition, it is clear there must be certain ways of making data governance a standard way of doing business.
Accordingly, below are enumerated eight best practices for data governance, to help companies begin to deliver it:
1 Tackle data quality issues at source. The further downstream a problem manifests, the more it costs to fix - and the corollary is that the earlier it is trapped, the less it costs. And inaccurate data is a pervasive, persistent problem. Especially in a country like South Africa, with its multiple languages and relatively poor schooling record, data quality is suspect at the point where it enters the organisation. This manifests in incorrect spelling of people's names, wrong address details, transposed figures in invoicing, and more. Downstream, these organisations have to address the errors with special software - it is far easier and lower cost to invest in error trapping software upfront in one place (at source), and to have double-checking of data captured, along with training of staff. This also calls for sound capture processes and operational systems that have data management rules fully entrenched and automated where possible.
2 Identify the source of the problem. It is not always going to be data quality, though, that captures business's attention: it's more likely to be a downstream issue, such as the cost of postal returns because of poor data capture processes. Any reduction on postal returns improves the bottom line - therefore improvements in data quality can also boost cash flow.
3 Consciously build data governance within your organisation. Data governance does not just happen, just as corporate governance and IT governance don't. It needs to be made part of corporate strategy, deliberately planned and delivered. And it needs to be actively advocated with passion, and its value marketed. No technology initiative has succeeded, anywhere, without this approach.
4 Appoint a data steward. Ensure the role and importance of it are acknowledged. Without stewardship of data, data governance will simply wither and die. The data steward sets data quality levels, calculates the cost (including the opportunity cost) of poor data quality, sets access controls, creates metadata capability and mediates inevitable conflicts, to name a few priorities.
5 Establish a data governance committee. Ensure it has teeth. This will be the central policy-setting body. This committee will have responsibility for creating a data governance process, infrastructure, and setting and enforcing policies and procedures.
It is clear there must be certain ways of making data governance a standard way of doing business.
Mervyn Mooi is director of Knowledge Integration Dynamics
6 Create and implement proper controls. At heart, governance is about control, checks and balances, all applied in such a way that the business knows that all decisions have been taken correctly, with accountability, traceability and auditability front of mind. This will support the organisation's integration competency centre (ICC) and/or centre of excellence for business intelligence.
7 Transparency must be the order of the day. The corporate intranet is ideal for publishing all background processes and all decisions taken. Avoid implementation or adoption of isolated, temporary, secretive or self-serving programmes. Standards-based decisions and processes dealing with data and data governance should be adopted and communicated to the correct audience as a company standard, which should also enforce transparency.
8 Ensure metadata is fully entrenched. Metadata is data about data - analogous to the listings on a CD or DVD casing. Metadata establishes naming conventions, data lineage and the projected impact of any changes to data. Without metadata, in fact, there can be no data governance, so make this one of the highest priorities.
Data has always been, is now acknowledged as, and will always be one of the most important assets for any company operating in the 'Information Age' - and that's most companies. Only with data governance will companies be able to bring it to heel and allow it to deliver its full potential.
* Mervyn Mooi is director of Knowledge Integration Dynamics.
Share