• Home
  • >
  • Editorial
  • >
  • DataIQ Member’s Briefing – Can we fix it? The problem with data quality

DataIQ Member’s Briefing – Can we fix it? The problem with data quality

DataIQ members agreed that quality is a must to become a mature and successful data organisation, but how can data quality issues be addressed?
Data Quality Pic 2.jpg

Know your journey 

“Don’t believe everything you read on LinkedIn,” said one DataIQ member from the HR sector. “Everyone is up against it when it comes to data and data quality.” It can be easy to think you are the only one facing data quality issues as LinkedIn becomes more saturated with feel-good stories and images of success, but this is far from the truth – rarely do people ever share their struggles and pitfalls on social media, particularly LinkedIn. When was the last time you saw a post detailing legacy data issues or digitalising data from books? 

The first step is to understand how and why data quality issues are arising within your organisation – knowing the problem is half the battle. The same member explained how they discovered their data quality issues were rooted in product ownership and engineering: “We’re very reliant on individuals following processes – working with a jungle of various systems to generate that data. We suffer from data quality issues often due to incompleteness, inaccuracy or unavailability.” This raises the problems caused by manual data generation, but it is often the only option available as automated data functions can be costly or simply not exist yet.  

Furthermore, many areas of businesses do not realise that data quality is even an issue until it is too late, which places data offices ahead of the curve and in a position to lead by example. It is down to the data team to instil knowledge and awareness of data quality to the wider organisation. This can be achieved through the development of a strong data culture and improved emphasis on why quality is such a determining factor for data benefits. “There is a lot that we can do as data leaders to get that culture in place,” added one participant. “We need to get people understanding why it’s important for them to work with data, and that it’s their responsibility to look for quality.” Another member continued: “This is about culture change, and it is probably the thing that humans are worst at. Also, it’s also not necessarily native to data leaders to think about culture all the time.” 

Create a narrative 

As is often the main hurdle faced by data teams, improving data quality usually requires investment and time, which businesses are reluctant to pay for. Therefore, CDOs need to be able to create a story that is clear and concise as to why data quality is needed and how it can be achieved. “People need to be excited about data and know its value,” added one participant from the IT architecture side of their business. 

One roundtable member told the group how they focused on supporting wider business goals from the data function and this led to reporting lines from finance which did not previously exist. From this, they were able to quantify the risk to contract renewals or failures to win bids due to data quality issues and demonstrate a clear ROI, from de-risking through investment in staff. Ultimately, this member crafted a narrative that they could de-risk the top line in the UK by tens of millions of pounds per year, with an ROI exceeding 1,000%. This naturally created a strong argument to take to the business leaders and secured investment for the data team.  

Additionally, CDOs need to be realistic with their stories. One participant described how their ongoing data quality programme was likely to take five years to reach completion due to the scale of the challenge ahead. Five years can seem like a long time, but if that is the realistic timeline then that is the story that needs to be told. The whole point of data is accuracy and insight, so CDOs must ensure this mantra remains with their storytelling, even when it is difficult.  

The business landscape is always evolving, and customer demands are always changing, which is what has been experienced by one member in the entertainment industry – also on a five-year programme – and data quality is essential for their next phase. “Core to this five-year development is aligning our data to some type of value case within the organisation,” they said. “Through the use of our historic data and the organic data we are now collecting, we can look at either driving incremental revenue, protecting revenue or increasing our inventory to sell more adverts.” 

Fixes for data quality 

One member discussed how their business had previously been prioritising short-term fixes to data quality issues based on the organic storage of their data. “We’ve had analysts from my team building pipelines and building their own tables on various data lakes across the organisation,” said the member. “A part of our transformation is to set up a data mesh architecture and create other products that the business can use to access and maintain quality data.”  

One roundtable member from the insurance sector highlighted how their business operates with a “face-to-face model between the customers and the agents, so this can mask some of the data quality issues that may arise about the customer.” It can be a challenge to record data from an in-person conversation or individuals, but it is not impossible. It requires a strong data culture and active data input from the organisation. This can come about in its most simple form – note-taking during the meeting – but it still flirts with the issue of manually input errors and a lack of consistency from employee to employee. These reasons are why a comprehensive understanding of data literacy and data quality benefits are needed across the board.  

A secondary solution to this data quality problem is providing a platform where customers can control and update their own data. This solution does itself bring up new problems such as outdated data information being used, incomplete data profiles and customers not appreciating the need of high-quality data., but it can eliminate employee errors and removes the burden of data input from their shoulders. 

There was a detailed explanation from a participant in the entertainment industry about their data quality strategy that has three key areas of focus: 

  • Exploring long-term foundational fixes, not short-term fixes.
  • Fix data as close to the source as possible.
  • Identify data quality issues that unlock the value case to achieve business objectives.   

These three points can be adopted by all types of organisations. With these three focuses, it is hoped that the quality of data will forever be improved, engrained into the culture of the business and actively work to allow different teams to achieve their goals.  

The participants of the roundtable ended discussions feeling more confident in their data quality journeys and the knowledge that they are not alone in facing this problem. Data quality can be a daunting problem as it can interfere with all aspects of business growth, but it is a problem that can be addressed and fixed with the right approach.

Upcoming Events

No event found!
Prev Next