The term “data governance”, which has been around for more than a decade, has become very popular in recent years. I do not know when, the river’s lake flow out: “digital transformation, governance first”.

As a result, we see: not only is a traditional provide, BI, master data management, data warehouse metadata management, data integration and data service software vendor in the said data governance, ali tencent and other Internet companies, large state-owned enterprises are also talking about data management, many enterprises are the data governance as a necessary measure of digital strategy, on the enterprise’s strategic plan of action.

In many enterprises and individuals who talk about data governance, I find that there is a general consensus on data governance: “Data governance is easier said than done”!

First, why do we need to do data governance, really think through?

In the process of data governance consulting, we often encounter the following dialogue scenarios:

Excuse me, why do you want to do data governance?

We need to establish data standards, improve data quality, and achieve unified management of data assets.

Then ask: Why establish data standards and improve data quality, and what if you don’t?

Common answers: There are many data quality problems, and accurate data reports cannot be provided, which affects business efficiency and cannot support the digital transformation of enterprises.

Again: What data reports and what businesses are affected?

Common answer: XX reports are inaccurate, statistical caliber is inconsistent, data islands between systems, and data integration is difficult…… Blah blah blah…

Then ask: why does it cause inaccurate data reports, inconsistent caliber and difficult system integration?

Common answer: The data quality of the data source is poor because the data standards are consistent.

We carefully analyzed such research results are floating on the surface, around the problem of data in situ, did not really think through why to do data governance.

Therefore, the first step of data governance is not to analyze data problems, but to analyze business problems, find the core business appeal of the enterprise, and define the goals and scope of data governance.

Second, data governance is not a lofty thing, is basically dirty work, back-breaking work!

Data governance is fire, the DAMA data management body of knowledge in the guide, data governance is located in the “wheel” middle data management, data structure, data modeling, data storage, data security, data quality and metadata management, master data management and so on ten big sum in the field of data management, data management activities provide overall guidance strategy.

When it comes to data governance, we often talk about it as a combination of enterprise strategy, organizational structure, data standards, management practices, data culture, and technology tools. Those who have no experience in data governance are bound to think: Wow, data governance is “superior”! It’s strategic, it’s standard, it’s cultural.

However, only if you have really done data governance people know: data governance is not only hard work, tiring work, but also a thankless, often carry the blame, the leadership can not see the value of the work.

Data governance is sometimes not understood. Data governance is a foundation project. What people see is always a “high-rise building” of data application. Data governance teams are busy every day, and the leaders do not know what the “group of people” are doing. However, whenever there is a problem with data, the data governance team is the first to be held accountable.

Third, do data governance, why the data quality is still poor, what to do?

Our company did data governance two years ago, and established a data governance platform, with metadata management, data quality management and other functions. However, there are still many data quality problems in our company, leading to almost no one using the BI system built. May I ask if there is a good way to solve them?

I have no answer to this question. The reason is that the data quality is poor and [BI](https://www.finebi.com/?utm_source=media&utm_medium=jj) cannot be used. Although this problem is common, the reasons are different in 9 out of 10 enterprises that have the same problem. You don’t want to give advice without doing a lot of research and knowing the context.

After data governance, can the data quality of the enterprise be improved? The original intention is to ask why data quality problems exist after access to a data governance system or implementation of a data governance project.

The problem is complicated. As in project-based data governance above, it is a matter of addressing the symptoms rather than the root causes.

Some enterprises think that data governance is a set of powerful data governance platform, as long as the platform is powerful, can manage the data, this is precisely another mistake – only tools, do not know that the nature of data governance is to manage data, rather than management procedures, scripts and tasks.

In addition, there are a lot of enterprises have data problems and have a great impact on the business before governance — passive governance, lost the initiative of data governance, often solve a problem leads to more problems.

What is the way of data governance and how to do it?

Data governance needs system construction: to give full play to the value of data, it needs to meet three elements: reasonable platform architecture, perfect governance services, and systematic operation means.

Data governance is not a quick fix. It is a long and ongoing process, with no magic bullet or quick fix. Only by turning data governance into a regular mechanism, just like when we eat and sleep every day, can we form a habit and a culture, persevere, stay true to our original aspiration, and make unremitting efforts, can we achieve the desired goals.