A commitment to first-rate data quality management is more than just good business at one U.K.-based gas utility
-- it’s a matter of public safety.
Wales & West Utilities is a privately owned yet heavily regulated gas distributor serving about 7.4 million people in Wales and South West England. Founded in 2005 when National Grid Transco sold off some interests in the area, Wales & West is executing a 30-year plan to replace more than 21,000 miles of metallic gas mains and distribution conduits with longer-lasting plastic pipelines.
The company decides which pipes need to be replaced most urgently based on industry-standard safety considerations related to age, composition and maintenance history. And the entire process depends on high levels of data quality, according to Martin Malin, business applications manager at Wales & West.
“We’ve obviously got a large number of assets, and [much of the information] on those assets is historical data that was migrated from the old system into our new systems in 2005,” Malin said. “The importance of the accuracy of that data is paramount because it has safety implications, particularly in relation to pipe replacement.”
A three-tiered approach to data quality management
Wales & West’s IT operations include approximately 120 servers. About half of those are used for disaster recovery and nonproduction testing, according to Malin. The company’s main business applications include hosted instances of SAP R/3 4.7 for enterprise resource planning (ERP), SAP Supplier Relationship Management, SAP Customer Relationship Management and SAP NetWeaver Business Warehouse for data warehousing and reporting.
Prevention is better than cure. Fix it at the front end if you can.
Martin Malin, business applications manager, Wales & West Utilities
Since its founding, Wales & West has taken a three-tiered approach to ensuring the quality of any new data generated by the company’s approximately 1,200 employees. The first part of the process involves validating data when it comes through the door.
“Prevention is better than cure,” Malin said. “Fix it at the front end if you can.”
But if questionable information about assets, customers, finances or anything else makes it past the front-end validation process, it might not get by the SAP software. Based on preset rules, the SAP ERP software can reject anything that looks like an anomaly. When that happens, Malin explained, a Wales & West error management team is deployed to examine the situation and determine if a data quality problem truly exists.
“It’s always a balancing act between getting those validation rules tight enough to ensure high-quality data without making [the systems] restrictive or [unable to cope] with exceptions and minor variations that are perfectly acceptable,” Malin said.
Malin said the third stage of the data quality management process involves a combination of data profiling and exception reporting to help identify any anomalies that may have made it through the first two stages.
New data quality management software completes the strategy
While Wales & West has strong processes in place to ensure the quality of newly generated information, Malin said it was apparent early on that problems with historical data persisted.
The historical data quality problems became clear when longtime gas industry veterans working for the company noticed slight discrepancies between company reports and their own extensive knowledge of the area.
“Certain individuals were spotting anomalies,” Malin said.
To combat the problems, Malin and his team took on the task of manually identifying data quality issues and cleansing any files that weren’t accurate. But it would be difficult for Malin’s modest nine-person IT staff to keep up those processes over time. The team ultimately decided help would be needed in the form of data quality management software.
“Getting this right on a one-off was one thing,” Malin said. “But we really needed to be able to measure this and manage it, and demonstrate to the regulator and safety authorities that we know exactly how good our data is and that we are improving the quality of that data over time.”
In early 2010, Wales & West began evaluating data quality management software options from SAP BusinessObjects, Informatica, Information Builders and Trillium Software.
“An obvious route would have been BusinessObjects, clearly, because of the SAP connection,” Malin said. “But I was conscience that it was essentially a BI platform, rather than having a specific data quality focus -- although they have a profiler and they have data quality tools.”
All of the products evaluated were capable of similar tasks, but Wales & West liked the way Trillium handled those tasks most. Malin’s team gave Trillium high marks for ease of use and its ability to work well with the company’s data quality validation rules.
“The way in which Trillium constructs those rules is quite intuitive and reasonably similar to [Microsoft Excel], which is a tool set that most of these guys in the business were familiar with,” Malin said.
Wales & West recently implemented Trillium and is defining business rules within the software that will help support the pipe replacement project. Malin estimates that the company has approximately 120 business rules defined thus far.
Looking ahead, Malin said he’d like to see Trillium focus on further enhancing its capabilities related to units of measure and asset management in general.
Wales & West eventually plans to expand Trillium’s focus from gas mains and distribution pipes to other assets within the company. But Malin said he also intends to use the software in non-asset-related areas. For example, the procurement department may eventually use the software to seek out duplicate vendor records. The finance department, meanwhile, will use it to look for anomalies or suspicious patterns in the company’s general ledger.
Data quality management advice from the trenches
Malin and his crew were fortunate to have superiors who immediately saw the safety and compliance benefits that the data quality management software project would bring. But proving the business case for data quality initiatives isn’t quite as easy for many IT professionals, according to experts and user interviews.
A seasoned IT veteran, Malin had some words of advice for IT professionals seeking executive buy-in for data quality projects.
“It’s absolutely vital that [senior management] recognizes the importance of data quality as a topic area and as a big issue for the company as a whole,” he said. “If you dig deep enough, there is almost certainly a financial business case.”