How to measure data quality
Perfect data is unlikely, but there are some fundamentals you must still measure
With data quality, as with most other things in life, you get out what you put in. If you can collect clean, comprehensive data, then your database will be bursting with insights that are just waiting to be mined.
So, you might think, all you need do to turn your data into gold is make sure you collect only the purest, most error-free data sets available. If only it was that simple.
Organisations may typically store data into orderly columns and rows (or both), but once you widen your net to include information gleaned from the web and social media, the variety of data-types make it more difficult to store, cleanse and find any analytical insights. The structured view then starts to disappear.
Capturing data is only worthwhile if that data is actually adding value. Whilst we can say with certainty that more data is being generated and collected by organisations as the 21st century progresses Forbes estimates that 2.5 quintillion bytes of data are created every day this is actually having a detrimental effect on the overall quality of data gathered. Simply, there is too much to handle.
Of course, if you want incredibly high-quality data, you'll have to spend money and time cleaning it. But if you're analysing customer sentiment on Twitter, for instance, and need to react quickly, you'll have to sacrifice quality for speed, or you'll potentially end up with upset customers.
All of this means that, in practice, perfect data quality is a nigh on impossible aim. The data you collect from various sources will be unstructured, and cleaning it costs. However, that doesn't mean you shouldn't value the quality of the data you hold. While it won't be perfect, you want to ensure it's as clean as possible, so that it remains useful.
When equipped with the key metrics of measuring data quality, enterprises know where they stand. Next would be to deploy a data quality management strategy, a process that further improves the measuring of data quality through applying the combination of the right people, processes and technologies.
So, how do I measure data quality?
There are a variety of definitions, but data quality is generally measured against a set of criteria called 'data quality dimensions' that assess the health of the data, such as completeness, or uniqueness.
In an ideal world, all these criteria would hold equal weight but depending on what you intend to use your data for, or its primary function, you may want to prioritise certain criteria more strongly than others.
Although many industries will have devised separate metrics for assessing data quality, DAMA International, the not-for-profit data resource management body, has set out its six key criteria that it considers as the standard for measuring any database against.
Completeness is defined by DAMA as how much of a data set is populated, as opposed to being left blank. For instance, a survey would be 70% complete if it is completed by 70% of people. To ensure completeness, all data sets and data items must be recorded.
This metric assesses how unique a data entry is, and whether it is duplicated anywhere else within your database. Uniqueness is ensured when the piece of data has only been recorded once. If there is no single view, you may have to dedupe it.
How recent is your data? This essential aspect of the DAMA criteria assesses how useful or relevant your data may be based on its age. Naturally, if an entry is dated, for instance, by 12 months, the scope for dramatic changes in the interim may render the data useless. Car mileage, which changes frequently, is a good example.
Simply put, does the data you've recorded reflect what type of data you set out to record? So if you ask for somebody to enter their phone number into a form, and they type 'sjdhsjdshsj', that data isn't valid, because it isn't a phone number - the data doesn't match the description of the type of data it should be.
Accuracy determines whether the information you hold is correct or not, and isn't to be confused with validity, a measure of whether the data is actually the type you wanted.
For anyone trying to analyse data, consistency is a fundamental consideration. Basically, you need to ensure you can compare data across data sets and media (whether it's on paper, on a computer file, or in a database) - is it all recorded in the same way, allowing you to compare the data and treat it as a whole?
Remember that your data is rarely going to be perfect, and that you have to juggle managing your data quality with actually using the data - spend too long on ensuring quality, and there'll soon be no point analysing it, because it'll be far past its sell-by date.
However, you should perform regular data quality audits - especially as you're probably regularly collecting new data sets - to ensure it's as clean and useful as you can make it. Without good data, you can't rely on it to produce useful business insights and to inform good decisions.
Why measuring data quality is important
Quality data can be the difference between enterprises keeping their heads above water, and sinking. This is particularly apparent when considering competitive markets, which are typically flooded with SMBs struggling to steal slivers from giant corporations. With rivals taking advantage of data and budgets already stretched to breaking point, organisations that aren't capitalising on the opportunities strong data can provide risk being left behind.
From a purely economic perspective, as data quality is optimised so are company finances. That's because poor data needs resources to transform it into insight. Research conducted by Gartner found that organisations believe they lose an average of $15 million per year in losses related to poor quality data. Having a data strategy in place would ensure a certain quality level will be maintained, reducing these outlays.
Accurate data also allows enterprises to better understand their customers' needs. This makes for more effective marketing, with targeted campaigns reaching the desired demographics. Internal processes should improve, as when decision makers are able to thoroughly trust the data they rely upon, better decisions can be made faster.
Companies also need to be aware of compliance regulations. In many industries, the process of storing data encroaches upon data-protection laws. The data must be protected to a standard, and must not be used for untoward purposes. With a better understanding of the data you possess, there is less chance of accidentally using data in ways that are restricted.
Digitally perfecting the supply chain
How new technologies are being leveraged to transform the manufacturing supply chainDownload now
Three keys to maximise application migration and modernisation success
Harness the benefits that modernised applications can offerDownload now
Your enterprise cloud solutions guide
Infrastructure designed to meet your company's IT needs for next-generation cloud applicationsDownload now
The 3 approaches of Breach and Attack Simulation technologies
A guide to the nuances of BAS, helping you stay one step ahead of cyber criminalsDownload now