3 Easy Ways To Give Your Data A Check-Up

Being able to trust your data is critically important to every business, especially when even the smallest slip-up in data quality can cause big problems further down the line. Getting your data booked in for a check-up, therefore, is just as important!

3 Easy Ways To Give Your Data Quality A Check-Up

In the same way that you know when you’re not feeling totally on top of your game, your data has tell-tale signs that not everything is working exactly as it ought. So, without further ado, here are three easy ways to perform a data quality check up, and move from feeling as though something’s not right to pinpointing exactly what you need to do.

1) Find and fix the weeds  

Any gardener can tell you that for good crops to grow, you need to keep on top of the weeds. After all, if you let the weeds take root, it can lead to loss of the whole crop. Likewise, even the smallest discrepancies, inaccuracies, or duplicates can throw your data off balance. This means that re-evaluating potential weaknesses and seeking to correct them is key.

Data Checkup: Find and Fix the Weeds
Datactics HQ Rose Garden in full bloom

But in just the same way as crops and weeds aren’t necessarily easily distinguishable without some green-fingered expertise, you need to involve the people who know what good looks like to address a data quality challenge.

Tom Redman, the “Data Doc”, has a super-handy method for figuring out just how big the data quality problem is. Head on over and have a read here, pick a Friday, schedule a Zoom call, open a beer or two and get cracking!

Once you’ve got your business teams to assess the theoretical size of the problem, you’re already in better shape. A data checkup will help you figure out whether improving the data could be achieved by removing data that isn’t useful, or filling gaps where data is limited, or making sure that your reports are fine-tuned on the problematic data elements, business areas or teams.

2) Talk to your front line  

Ultimately, the people who deal with your data day-in, day-out, are the ones at the coal face, the front line, who are capturing data at its source and updating data records at a phenomenal rate.

Talk to your front line
What do we want? Data!

If you were to poll your people on the quality of data, and whether they understand who is responsible for data quality testing, what would the result be? If it doesn’t bear thinking about, then that is pretty much the answer to your question! Find out how you can empower people in your business process to be as engaged in the data quality story as they are in analytics, winning business and managing customer experience (to name but three). A good starting point is to look at your organisations data quality management processes, including data governance. Within this, data stewards can perform tasks such as data profiling and data monitoring using the data quality dimensions of accuracy, timeliness etc. This will be a useful benchmark for members of the data team to measure the quality of their data long term.

Often, accurate data is a result of trained and competent employees. However, the ever-changing nature of data and the increasing rate of regulations has meant that a manual approach isn’t enough anymore to achieve high quality data. However, it is a good start for the data team to analyse the data at hand. Recognising there is a data quality issue is the first step towards giving your data a health check. Once you have established there is an issue, a data quality solution can be sourced to solve the problem.  

3) Talk to your customers!   

Talk to your customers!
I find your lack of customer service disturbing

Of course, quite apart from your internal discussions on whether things are in a good place or not, there’s nothing quite like hearing it from people who rely on you to manage their customer data.

Ask how many poor customer experiences resulted from bad data, or processes that didn’t align with how the information was to be used. Find those situations where a customer was contacted despite actually having died, or a policy was closed without the customer being aware.

Conducting a “five whys” process into poor experiences is a great start. Ask “why was this a result of poor data quality?” five times, coming up with five different answers, and follow those rabbit holes to the source of the problem. Focus on the problems that matter most to customers and build a comprehensive business case that demonstrates just how customer-centric you really are (and what to do next!). In the long term, this will help establish data integrity within your business.

picture of Matt Flenley
Matt Flenley
Marketing Insights

To find out more about what we offer in the data quality and machine learning fields, drop me an email or connect with me on LinkedIn and let’s continue the conversation!

Scroll to Top