Tim Hoyland, a partner in Oliver Wyman’s aviation practice, discusses how distinguishing good quality data from bad is crucial to MROs looking to enhance their competitive edge.
When migrating to new IT systems, airlines must face the sins of their past.
Years of poor data collection practices or weak discipline among employees about recording data have resulted for many carriers in data that is incomplete, inconsistent, or invalid. Poor data integrity has significant short- and long-term business impacts, as it is much more difficult and expensive for staff to find reliable data to make day-to-day decisions.
For maintenance, repair, and overhaul operations, data can be crucial for offering efficient service. Poor data about, for example, parts availability, can cause costly delays.
Many of these impacts are hardly visible to management, making it difficult to measure the effect of poor data integrity and to address the problem. But consider that new IT systems improve organizational effectiveness and safeguard regulatory compliance for airlines and the maintenance, repair, and overhaul industry. The same is true for investment in clean data.
What is data integrity?
The terminology used for data cleanliness can be confusing. There are two mutually exclusive aspects of clean data: quality and integrity.
Data quality refers to the accuracy and clarity of the meaning, context, and intent of data. The quality of data affects the ability to create reliable and accurate knowledge that leads to greater insights. Even though data might meet all criteria for integrity, it still might lead to inaccurate knowledge if the quality is poor.
Data integrity refers to the validity, completeness, and consistency of the data. Does the data conform to pre-established business rules? Are there missing or erroneous values? Is the data consistent across all databases? Identification of data discrepancies is straightforward, but data cleansing can be laborious. Once cleaned, data integrity can be maintained by built-in business rules and relational databases. Shortfalls in data integrity make up the majority of the types of discrepancies in IT systems. For the scope of this study, we will focus on data integrity.
Significance of Data Integrity
Data integrity directly impacts operational efficiency. While cleaning up the data might be laborious and expensive, operating with bad data is even more costly.
Employees have to spend more time managing systems with poor data. As people lose trust in the data, they use it less often, and miss out on the benefits the data systems were meant to provide. The more complex and erroneous the system, the more people are needed to understand the data, and the more work required to put out fires that arise from bad data.
For example, incorrect information about parts positioning can cause operational delays for aircraft on ground situations. Persistently bad data could lead to local employees stocking more parts, and could create more work for employees to locate parts when needed, two issues that data are supposed to solve. This can result in longer ground time for aircraft.
Bad data also hinders the audit process; it’s much easier to provide information to regulators from a clean data system. Bad data interferes with configuration management and maintenance schedules, and can complicate overflies, parts loans, tooling calibration, and parts shelf life.
Still, cleaning up data is expensive, as the volume of data can be enormous, and the intervention must take place on a live system that people rely on for operations. Further, the airline has based so much of its operations, sourcing, warranties, repairs, and other systems on bad data, that cleaning up the data can have broad impacts. These data issues can also increase the cost of IT system migration, creating roadblocks for change, adding time and cost to the process.
Sources of Bad Data
To address poor data integrity, it is important to understand common sources of bad data. In the aviation industry, data errors typically result from the complexity of the system, lack of system controls, and a culture of neglect.
The industry’s inherent operational, revenue, and regulatory complexity forced carriers to hold multiple levels of data in multiple databases. Most of these databases are connected through complicated mapping landscapes that are only understood by a handful of people. Additionally, the large volume of data that must be tracked has increased the complexity of the system. Few companies are able to handle the velocity of change, forcing carriers to rely on temporary solutions that can lead to more bad data.
Lack of operational controls, such as manual entries, free text boxes, and lack of automation, also leads to poor data integrity. If a system requires a manual entry, human generated errors are inevitable. If a system does not include the correct rules to appropriately structure a data entry, an error could occur. Manually inputting entries tends to lead to more unclean data entries than automated data entries. Further, if a data entry is inputted inaccurately the system does not automatically correct the error or notify the system user.
Organisational culture can also lead to poor data integrity. Many employees do not understand of the impact of discrepant data on business operations. It is often easier to neglect procedures in order to finish tasks quickly. Further, some employees think IT will clean each data entry, but the IT department generally lacks the time or business knowledge to do so.
Steps to improve and maintain data integrity
The key to maintaining data integrity is to make it part of the company’s culture. Leaders must take the problem seriously and take steps to change organizational behavior about data integrity.
Improve existing business procedures to ensure completeness, accuracy, and consistency of data across all systems and databases. Give employees the right tools to enter good data, develop and standardize the cleansing process, and build an accountability structure. If it does not already exist, establish validation points that identify and reject mutually exclusive combinations that could be entered unintentionally by users.
Make clean data a long-term goal for the company. Establish standards to maintain cleansed data. Proper data monitoring and audit techniques are required to achieve data integrity.
Continuous data quality management has been a key focus for other industries that are data rich and operationally heavy. Most IT system providers have refined their product offerings to include point-of-entry validations, database cross-referencing, and build-in data quality alerts. But the aviation industry has only recently begun discussing the problem, and most long term data cleansing programs have been focused on customer and revenue data.
Little has been done to build sustainable programs within the MRO sector. At a time when some MROs are struggling to find long-term relevancy, data integrity could be a valuable asset to clients.
Tim Hoyland is a partner in Oliver Wyman’s aviation practice, and he is based in Dallas. Konstantinos Varsos is a principal in the aviation practice, and he is based in Boston.