What NOT to do for improving data quality

Repost Linkedin : https://www.linkedin.com/posts/yacine-bekka_dataquality-datamanagement-digital-activity-7034069227783249920-JuHr?utm_source=share&utm_medium=member_desktop

---

Hey, it is well-known that data quality is not a big deal, right… ?

So let’s look at 6 tips together to make sure it doesn’t improve :

1. Treat data quality as a technical problem only

• Use only technical solutions to improve it (e.g., software, probabilistic controls, etc.).

• Do not involve business SMEs (data producers, data consumers) to ensure that data quality management is as far from business priorities as possible and that they do not understand it.

2. Do not use common and clear guidelines for data quality management

• Stress everyone in the company about bad data quality and that it’s their job to improve it.

• Do not define any systematic process for data quality management or any guidelines.

3. Focus only on theoritical aspects

• Make sure to have all data quality dimensions quoted in the literature (60+) covered by your data quality process.

• Make your process documentation highly complicated and full of jargon. This way, nobody will be able to contribute to it or understand the meanings and methodology of indicators.

4. Perform only data corrections, root cause don’t matters that much

• Focus all effort on data correction and firefighting quality issues where the “symptoms” occur.

• Forget about root cause analysis and a holistic view of why we have an issue. It’s completely unnecessary.

5. Treat all quality issue as the same no matter the business impacts

• The real business impact of quality issues is irrelevant; make sure not to prioritize the high impact ones.

• In fact, to save time, you can just not assess the impact of quality issues and use instead the “wet-finger” prioritization technique

6. Launch isolated data quality projects

• By avoiding a common and continuous approach, you ensure information loss and prevent follow-through of the issues previously identified.

• It also helps to create or reinforce existing silos in the company, which is always a good thing, especially when it comes to data.