Reasons for Data Governance Project Failure
Original

ZenTao ALM
2022-07-13 10:18:02
1743
ZenTao: 15 years of dedication to building open source project management software
Download Now


In the digital age, data as a new factor of production has received unprecedented attention from all walks of life. As more and more data becomes available, the question of managing and using it well and making it valuable has become a challenge for many companies. Data governance is then applied. Effective data governance ensures that enterprise data is comprehensive, consistent and trustworthy so that the value of data can be leveraged to improve business efficiency, enhance business growth opportunities and drive digital transformation.


Sounds easy enough, right? But the truth is that data governance is a big challenge for each enterprise. According to a Gartner survey, over 90% of data governance projects fail to perform well. Why do so many data governance projects fail? Those new to data governance may find this set of statistics a bit exaggerated, and Many would even be discouraged. But that's the reality, and I'm accustomed to it after years of work. Here's a summary of why data governance projects fail to perform well.

Lack of clear objectives

Many people who work for data governance lack clear objectives. It cannot be said that there is no objective, but the objective is set very large, very general, and unfocused without considering the achievability and measurability of the objective. For example, the objective is to solve all enterprise data quality problems. The data governance objectives are too short-sighted, leading to a rework of data governance. For example, the definition and understanding of data quality objectives were not agreed upon by those involved staff, and governance was implemented in the presence of disagreement. Data governance objectives are not linked to business objectives. It only takes a technical view of how to govern, just for governing.


Tips: The absence of clear objectives or a focus on short-sighted governance objectives without mechanisms for ongoing governance can lead to wasted resources, leading to data governance being set aside before it can make an impact. Effective data governance starts with a clear governance objective that should be tied to business value.

Confusing division of labour, unclear authority and responsibility

There is no clear definition of data ownership, access, management, etc. It is claimed that everyone is responsible for data quality. Still, data management is confusing, with a lot of duplication, and no one is willing to take responsibility when mistakes are made.

There is no precise mechanism for data accountability, no idea who to ask when data issues arise, and coordination among multiple parties leads to slower project implementation and results in many quality issues.

IT staff should focus on the definitions and trends of data quality and analyze the root causes of data quality issues. It's okay not to know the business. Just learn it. Business staff should analyze data structures and understand data linkages and data usage. It's okay not to know the technology. Just learn it.


Tips: Effective data governance requires data accountability and accountability, as well as collaboration between IT and business. While IT should focus on technology delivery, business needs to focus on defining data quality rules and improving data quality. Both teams should work together and maintain open lines of communication to monitor and improve data quality.

Lack of attention from senior management

Senior managers lack awareness of data governance, confuse data governance with data management, and believe that data governance is the matter of the IT department or DBA without too much involvement and attention from senior leaders.

Senior managers lack authority and influence to drive data governance objectives and business performance. Regarding cross-departmental coordination, each department promises to cooperate, but in practice, they do what they want, which causes a data governance strategy to be virtually useless.


Tips: A practical data governance project requires senior leadership to take responsibility. The leading senior managers should not only have a detailed understanding of data governance but also possess considerable authority and influence, coordinate across departments, and give sufficient authorization and strong support to the data department in the project.

Lack of data governance experts

Conflating data governance with systems management and making IT system administrators responsible for data quality. This is as unreliable as making the person who fixes the water mains responsible for tap water quality.

Many believe that data quality management is all about the IT staff and that knowing algorithms, models, and programming is all needed. What they don't realize is that the data quality team has to be able to make the right decisions with a business analysis mindset and a good understanding of business processes. If they don't understand the business, they may not be able to understand the impact of incorrect data.

Many people believe that data quality is all about the business staff and that it is enough for them to be responsible. They don't realize that data quality is not just about identifying business rules and correcting errors. It's also about continuously monitoring data and designing processes to minimize the risk of mistakes. What's more, in many companies, not all business staff can articulate the business rules.


Tips: Data governance is cross-functional, not a matter of a department or a person. Pure business personnel and isolated technical personnel do not have the complete ability to deliver data governance. Enterprises must cultivate a group of data governance experts who understand data governance technologies and enterprise business.

Opaque rules and systems

The formulated data management system and process are not released and made public, the defined data standards are not publicized, and the relevant stakeholders do not know whether these rules are clear. Anyway, we have finished our work.

Data governance progress and results are not reported in time, and the results can not be delivered to relevant leaders and departments. We are the real "data craftsmen", busy processing data daily and debugging procedures, and have no time to engage in other things.

"Money can not be leaked", data is an asset, you can "hide" it well, there is an "information island" to maintain the "information difference" with other departments, to ensure our "mystery".


Tips: Effective data governance requires complete transparency. The project progress, the work results, and the problems need to be seen by the boss and the business in a timely manner to enhance their confidence in data governance. Problems should not be hidden, and they should be exposed in time and be timely solved. At the data level, it is also necessary to be more transparent. Master and reference data should be shared across the company, and data assets should be visualized as much as possible.

Passive data governance

We should focus on business processes rather than data quality, and data quality becomes an issue when it leads to poor decisions.

Many do not consider the proactive strategy for establishing data governance or holding uniform data standards. Still, with data maintained separately for each system, data quality becomes an issue when systems cannot be integrated effectively.

Usually, people pay no attention to data governance and timely handling of data quality issues. Then data quality becomes an issue when the regulator issues a fine.


Tips: Effective data governance requires a data governance strategy at three levels: before, during, and after. Before: defining and establishing data standards, communicating and training on data standards, and fostering an enterprise data culture. During: data validation based on data standards, data maintenance and use based on established processes and systems. After: continuous data quality measurement, continuous data issues, business process improvement, etc.

Project-based data governance

Many people viewed data governance as a one-off project, and expectations were initially high. They believe that data quality would improve overnight with the implementation of a project.

Data governance is all about dealing with current data issues. There is no need for us to define rules or write documentation.

Data quality and data governance processes are a single, one-time activity. There is no need for us to establish a continuous mechanism.

Data governance strategies and data quality measures do not need to reach an agreement with the relevant departments. We need to complete the project task.


Tips: The ultimate goal of data governance is to improve the value of data. It is a long and continuous operation process that needs to be gradually improved and iterated. It is unrealistic to expect data governance to be completed in one step. Project-based data governance is not comprehensive and has no continuity. It can solve temporary data issues, but it is challenging to obtain sustained data value, and the effect is unsatisfactory.

Isolated data governance

Data standards are established but not updated, and old systems are not transformed with data. New systems are built without reference to data standards; thus, data standards become useless.

We should treat data governance as a separate and additional task do not tie it to business processes. The business department only carries out the clean-up of data quality issues but does not accept adding data rules to the business process.

Business apartments believe that data governance only adds to their workload and imposes constraints on the business without helping or adding value to their business performance.


Tips: Effective data governance should be viewed as a tool to help business staff achieve business goals, not as an additional task, but as embedded in the business processes of the enterprise, regulating the maintenance and use of data in the daily life of the business.

Instrumentalism

Build data standards? Haven't we already purchased a data governance platform? Why is there no data standard in the platform?

Collecting and correcting meta-data? I remember that our data governance platform can adapt to dozens of database types, so can we collect whatever data we want?

Is there still a problem with data quality? Is our data governance platform not functional? Should we purchase a new one?


Tips: Instrumentalism, too much emphasis on tools and technology, neglecting the construction of data governance organization, culture, systems, processes, standards, and other systems. The essence of data governance is to manage data, but it has gone astray in managing procedures, scripts, and tasks, resulting in a loss of focus in management.

Words at the end

Data governance is about "how to manage data" and involves a series of strategies, such as strategy, culture, systems, processes and standards, which are the core elements of data management. Many factors influence the development and implementation of each data governance strategy, which can lead to data governance failure.


In this article, we have shared various reasons for the failure of data governance projects. Although some of the entries are somewhat exaggerated, there have been similar cases in projects where individual factors may not mess up a data governance project by the minute, but certain factors combined may cause the project to fail.

Write a Comment
Comment will be posted after it is reviewed.