Data Governance and Stewardship is often a practice overlooked or executed in a way that's difficult or costly to maintain. Not having a solid practice can lead to legal issues, but the true cost to a data owner is often hard to figure out. Some common problems we see today include the following:
1. Overall poor data quality
Overall poor data quality can lead your employees, and processes, to take misguided paths. Let's say that during a sales opportunity, confusion over an existing customer address leads to the customer being entered twice in your pipeline database. Not only is the customer now treated as 'new', now you have two records for that customer with at least one being 'redundant'. When it's time to make a sale, the rep thinks the customer is new and offers more incentives.
2. Lack of data visibility
Lack of data visibility creates redundancy and makes integration projects both time consuming and costly. If your organization is large and your data landscape is not entirely visible and exposed to your IT teams, creating databases that contain the same or similar information can happen. Remember the product table that both the Sales and Warehouse management team implemented? Not only do you pay for duplicate storage, when discovered, you incur the cost related to integrate the systems. And when data is not accurate, the integration of these systems requires process, procedure - and perhaps compromise as you try to reach "one version of the truth".
3. Unclear ownership
Unclear ownership of data leads to (1) data integrity problems that are difficult to resolve, and (2) development initiatives that are hard to instantiate.
- Who owns the data you use?
- Are data issues your responsibility to fix?
- Should IT fix this bad address or should it be the business user?
- Who in IT owns this data?
With so many questions over ownership, it is unlikely that someone will step up and fix it. This means that bad data remains, and costs snowball over time. In larger organizations, getting funding and approval for IT projects can be a problem. If ownership is unclear, just starting the process will be a major headache!
4. True impacts unknown
When the true impacts of development initiatives are unknown, poor risk calculations and failed deployments are likely to follow. Let's say a user wants a simple change to the calculation of a key performance indicator, and the developer has no clear idea of the impact. It turns out that the field is used by 10 downstream tables and 100 reports. The change is made and over 1000 users are affected! Due to the low level nature of the change, it's not entirely clear what caused the user problems and the deployment is "rolled back".
5. Preventative action is hard to define or resource intensive
Preventative action is often hard to define or it involves intensive manual activities. You've been charged for returned mail often enough, and you think it's time to fix the problem. But how do you actually fix the problem? Is it just the zip code that is wrong, or it is more than that "simple" fix? If you’re not entirely sure which part of the address is wrong, it is difficult for IT to implement validation measures to avoid a recurrence. You could fix the address manually, but how long will that take? With over 1000 customers it could be a full time job!
6. No ability to quantify data issues financially
So you have plenty of data problems - but how do you decide if they’re really worth fixing? Are some departments suffering more than others? It doesn’t make sense to start a $10,000 development project if your bad address data is limited and can be resolved quickly. However, if your product data is bad and leads to misplaced orders that generate significant shipment returns, the cost can be huge. Cost of analysis for bad data can take time - and with no system of record, this time can be wasted. If these 6 reasons why companies gotta have data governance sound compelling to you, I hope you stay tuned as we go through a Data Governance and Information Stewardship series.
Photo Credit: Victor meldrew by red5standingby, on Flickr