Data quality as a driver for your success

Article

Data quality as a driver for your success

A short guide to successful data quality projects

An increasing number of companies are using their data for a variety of purposes; for instance to do analytics, gain insights or to meet higher standards of regulatory requirements. Nevertheless, when working with data one often comes across many data quality issues. It is well-known that ‘garbage in means garbage out’, meaning that data needs to be of high quality before we can actually derive quality output. This article is a first in a series of articles that explores how to get most out of your data – starting with data quality.

Your challenge

While data can be an asset in enabling efficiency improvements and cost reductions through automation and in helping business analysts provide accurate information for strategic decisions, it can be a liability too. Preparing an accurate finance and risk report requires good quality data, failing which a report could contain mistakes, thus possibly increasing costs and even creating a reputational risk. So, good quality data is crucial for a company in bringing trust in decision-making and in communication.

Our experience has shown that Data Quality improvement programs often fail to reach their full potential, for a variety of reasons:

  • Prior Data Quality initiatives are focused more on IT & data sources instead of on how businesses use data;
  • Instead of a strategic roadmap to drive sustainable change through culture, behavior, mechanisms and clear governance and control, Data Quality programs follow a solution-based approach;
  • ”Quality” is not clearly defined and may mean different things to different audiences;
  • Stakeholders’ expectations are not effectively considered and incorporated in the improvement initiatives

Why Deloitte

Deloitte has extensive experience with Data Quality programs and managing projects. We work in multidisciplinary teams with a balanced mix of technical skills, as well as soft-skills to enable us to handle such challenges. Navigating a complex stakeholder field is an important component, as experience shows us that this challenge is often overlooked in Data Quality programs.

Engaging your stakeholders is one of the most important facets of improving Data Quality in an organization, as Data Quality problems often cannot be fixed by project teams but have to be fixed by the data owners. By starting small and delivering a new or improved working product every two weeks, we encourage data owners to feel ownership and build trust by showing results.

Before starting such a Data Quality project, however, we have to be careful not to get ahead of ourselves. If a data landscape is immature a solution-based approach may not work, as you would be treating symptoms rather than the actual problem. We strive to make a lasting impact instead of delivering a one-time project to provide a temporary solution

Our solution

Once preliminary research shows that a Data Quality project is the right approach, in our experience the following factors are the most important for its success:

Ownership:

In a major project for a Dutch bank a number of people were working on maintaining ‘data flows’. A well-known problem throughout the project regarded the significant Data Quality issues, which meant a serious risk to the success of the project. We were asked to establish a Data Quality task force to combat these issues without the team having to deal with the daily routines and responsibilities that the original data squad had.

In avoiding the interruption of daily routines, we managed to split up long and complicated processes in ‘increments’ – small steps that we could define as a ‘working product’. We focused on the ownership of each product and consulted the owners to discuss exactly what improvement was needed. We subsequently automated these steps as quickly as possible – two weeks per ‘increment’ instead of three to six months for entire processes. This would not disturb the daily routines and offer the ‘product owners’ and anyone who worked with the product the opportunity to set priorities and keep close track of ‘delivering the right things’. This is how we integrated the governance into the project.

Teamwork:

Another success factor is working in multidisciplinary teams whose members have complementary skills and expertise on, e.g. robotics, analytics, agile working, assurance, data management and data governance, and learning how to combine these skills. The team members were able to challenge each other, review each other’s work, and become better at what they do. This was very rewarding and it helped to create an inspiring and fun atmosphere. Ownership and personal development were continuously encouraged, an important part of any transformation.

Of course, many other factors are important when improving Data Quality and no two issues are completely the same. Still, in repairing Data Quality issues in your organization, strengthening ownership and teamwork are both essential.

Other issues may often be at play that could prevent an organization from reaching their full Data Quality potential. The underlying problems causing Data Quality issues are an ineffective data infrastructure, data governance or data strategy. Our next blog will delve deeper into the details of these topics.

Did you find this useful?