Sometimes it feels like there are as many definitions of data quality as there are data points in Mestro Portal. But there are some basic pillars, starting points for working with data. Perhaps first and foremost that data is information and just as with everything else in life, there is both relevant and irrelevant information. Qualitative data can be described as something that serves its purpose in planning and decision-making. Qualitative can also describe data that closely represents reality. And not surprisingly, bad data = bad decisions.
Divide and conquer
In order to offer a more nuanced understanding of how we at Mestro relate to data quality, we have broken down the concept into its most relevant components – guidelines that in their own way define what quality actually is.
- Availability is for us often about uptime, a reference to the time our product is in operation and retrieving current data. A key figure for us is uptime per unit of time, how many times during a certain period we retrieve current data.
- Correctness refers to the fact that the data represents reality and/or that the data is consistent with the data source. We need to ensure that the data presented in Mestro Portal fully represents a data source such as a submeter. Unlike many others, we take into account different time zones, winter time / summer time and check monthly values against hourly values, all to ensure that the data is not misleading.
- Comparability means that our customers should be able to ask the same question in two different places and know that the answers are comparable. For us, this often involves different nodes and requires a lot of setup and tagging. It also requires some use of filters to ensure that different points are comparable.
- For us, completeness is about data coverage and our distaste for gaps in data collection.
- Clarity is key. It is difficult to act on the data without comprehensible visualizations and educational explanations.
- Reliability is the expectation that data is expected to be presented in the same way across the product. We want to be as close to the original source as possible, to minimize the risk of changes and inaccuracies when data passes through redundant stages.
- Transparency is a relative of clarity but we consider it of additional value as we need to be able to demonstrate how data stacks have been processed at different stages. Where does the data point come from and how has it been processed along the way?
- Flexibility means that data can be combined and used for different purposes, e.g. as a basis for further analysis in other software. One way to stay flexible is through the use of our API.
- Relevance is of utmost importance for all data. We need to have the right kind of data sources. This may mean that the data can be used to comply with certain standards and work to meet requirements and regulations.
- The data must also be reasonable. To ensure this, we continuously check and validate incoming data. We also offer alarms for deviations in data retrieval, all to ensure that the data is ultimately reasonable.
- Validity means that what you measure is actually relevant to what you want to do. Data should serve as a basis for meaningful KPIs.
- Last but not least, data should be verifiable. We must ALWAYS be able to check it against its original source.
The sum of its parts
It is difficult to place these perspectives in a hierarchical order, perhaps mainly because each in its own way plays an important role in the process of ensuring quality. And quality is more than just a goal, it is our service’s raison d’être. That is why we work continuously on this process.
For as much as the above list is a tool for valuing data, it is a filter through which data must pass. A data point may be relevant, but if it cannot be verified, it has no value for many purposes. In other words, the pursuit of high quality is as much about defining the goal as it is about defining the obstacles.
Mestro Portal is a service under constant development and this means that the way we define qualitative data must also evolve. Look at this as a postcard from a workplace where the work of ensuring data quality is the same as ensuring the value our service can offer.
And we will always offer high value.