It’s been said that there’s no such thing as “bad data” – that having any amount of numbers, figures and statistics to back up your business decisions is preferable to not having them at all. The problem is, data can be bad. It can be incomplete or outdated, calculated incorrectly or applied clumsily.
The Care Quality Commission (CQC) learned this lesson the hard way, after it found errors in the way it assessed the performance of general practitioners. John Flather, a Suffolk general practitioner whose practice was incorrectly labeled “high risk” by the CQC, told the BBC, “Our reputation, which has been built over many years, has been tarnished by incompetence that they purport to eradicate.”
Flather isn’t alone. Because of skewed assessments, the CQC misclassified about 60 GPs as “high-risk,” while hundreds of others are also expected to be reclassified.
This is the second time in a matter of weeks that an entity within the Department of Health has been found to have improperly managed data. An October report commissioned by NHS England found that “incomplete information” and “inaccurate data” were instrumental in the wrongful suspension of Leeds Teaching Hospital in 2013. That suspension only lasted two weeks, but the damage to Leeds had already been done.
The origin of both incidents – bad data – is more common than many realise and, what’s worse, businesses aren’t putting the necessarily fail-safes in place to prevent bad data. In an Information Difference survey, less than half of companies said they deployed data quality technology in their corporate systems, while seven in 10 said they didn’t measure the cost of poor data quality.
The silver lining to these figures is that there are plenty of examples of companies that have turned around their data management processes, through better data quality tools and new approaches. With that in mind, here are three basic steps to put your company’s data back on the right track:
1. Do the ‘Janitor Work’
How accurate is the information in your databases? Chances are, if your data management process requires extensive manual data entry and analysis, you’re probably housing more errors than you realise. Human error is often cited as the chief reason for data inaccuracy.
To offset these mistakes, it’s important to not overlook the “data janitor work” that’s involved in cleaning up data before it can be analysed. While the data cleansing process can be mundane, it will save you headaches down the road.
2. Break Down Data Silos
Another common data-related problem is that business departments are sometimes so siloed off that they don’t properly communicate and share information with each other, so their data sets are incomplete. This then leads to “insular thinking and decision-making” and only a partial view of the big picture.
As has been proven by companies that do data management right, ad hoc only adds problems. Some degree of data centralisation, and a willingness to bust silos, is needed.
3. Build an Infrastructure
While the first two steps toward eliminating bad data are beneficial in a vacuum, the only way they’ll have a lasting impact is if they’re actually incremental steps along the way to constructing an overarching data strategy.
Yet, many businesses don’t have a data strategy in place. This is ultimately a mistake. One solution to human error is to adopt technology and institute processes that minimise the role of employees in the data management equation. Whatever can be automated and regimented in the data management process, should be.
A Good Day to Eliminate Bad Data
It’s not easy for a business to rid itself of bad data – it’s too pervasive and can spring up at any moment. But, with the right data management strategy and infrastructure in place, you can reduce bad data and limit its effects, in turn saving your customers, your revenue and your reputation.
For more information on how to better manage your customer information, please click here.