- October 30, 2020
- Posted by: admin
Whether you’re an accomplished business manager or a consummate data wizard, Data Governance can be a lot to wrap your head around. Luckily, this list of common questions and straight-to-the-point answers will help you get a better grip on the basics.
Here’s our breakdown of the most frequently asked questions in Data Governance.
Data Governance is a strategic program for optimising the way a business deals with data. It aims to organise and improve the policies and procedures a company uses to define, collect, store, secure, manage, and monetise business data. Good Data Governance aims not only to avoid liability but to find new ways of generating value for a business.
Data Governance involves people, processes, and information technology. It evaluates and redefines roles and responsibilities, augments policies to improve communication and sharing between departments, defines and expands access to business-critical data, and standardises data collection and handling practices to ensure the quality and consistency of your company’s data.
Data Governance helps companies avoid liability, save time and money on bad data, improve customer relationships, and actively generate revenue. With accessible, well-defined and quality controlled data, organisations are 23 times more likely to acquire customers, 6 times as likely to retain customers and 19 times as likely to be profitable as a result.
Data Governance is a long-term strategic business program rather than a single short-term project. Implementing Data Governance requires making structural changes to a company’s existing data policies and practices, as well as redefining the roles and responsibilities of data handling personnel.
The best way to explain Data Governance’s importance to management is to focus on its benefits to the bottom line. Emphasise how Data Governance will advance corporate strategy and help achieve concrete business goals. It’s essential to communicate that Data Governance is as much about generating value as it is about avoiding liability.
The first step in implementing Data Governance is to develop a crystal-clear understanding of your company’s corporate strategy. Identify concrete business goals that can be achieved with data and define which specific data elements you’ll need to achieve them. With all this in mind, determine where improvements can be made to your existing data and procedures and begin drafting up new policies accordingly.
Measure the success of your Data Governance program by establishing key performance indicators (KPIs) ahead of time. Be sure to tie your KPIs to your organisation’s particular corporate strategy and concrete business objectives. This will ensure that you measure success in terms that are relatable to management and meaningful for the business as a whole. KPIs might track improvements in enterprise Data Quality, growth in the number of terms defined by business glossaries, or reductions in the amount of time spent searching for, organizing, and cleaning data. Another way to demonstrate success is by documenting improvements to your organization’s Data Maturity level over time.
Data Governance programs usually fall apart when they fail to secure business buy-in. It’s very important to have a business sponsor who can ensure that both business management and IT remain actively engaged with the program. Data Governance programs are also vulnerable to overreach. It’s important to establish sensible short term goals that are well-aligned with corporate strategy.
Data Governance deals broadly with organisational strategies, policies, and procedures. It provides executive oversight and dictates how data should be handled to advance business objectives. By contrast, Data Management deals with the tools and practices used to handle data and implement the policies outlined by Data Governance.
A Data Owner is an individual ultimately accountable for the quality of one or more data-sets. They are usually a senior-level employee equipped with the authority, budget and resources to define, clean and maintain the data they “own.” A Data Owner is usually not the same person responsible for managing the data day-to-day.
A Data Steward is an individual responsible for managing one or more data sets on a day to day basis. They report to a Data Owner and work to maintain the quality and security of the data. A Data Steward may or may not have any decision-making authority over their data.
A Data Owner is formally accountable for the quality of one or more data sets, whereas a Data Steward is responsible for the day to day management of the data sets themselves. In some organisations, the duties of both roles will be carried out by the same person. Larger organisations will assign these roles to multiple individuals to promote oversight and accountability, to reduce the workload on senior staff, and to encourage the participation of both IT and business departments in Data Governance.
Data Quality deals with whether or not a particular data set is fit for purpose. In other words, data that can be used to achieve a particular business objective is considered high quality. DAMA International measures Data Quality in terms 6 key dimensions: accuracy, completeness, consistency, timeliness, validity, and uniqueness.
Read More: Top 5 Data Quality Tools
Data Maturity deals with a company’s ability to collect, manage and monetise data. Data Maturity is focused on an organisation’s capacity to utilise and manage its data, rather than on the quality of the data itself. Measuring Data Maturity doesn’t need to be a complex process. It requires looking at a number of different organisational practices and data-touch points. Find out more here.
Data Lineage deals with tracking the complete lifecycle of a particular data set or element; where it comes from, where it’s transformed or stored, and ultimately how it’s put to use. Data Lineage is especially useful for determining the trustworthiness of your data, as well as for running impact and root cause analysis on data errors.
A Business Glossary defines the meaning, format and uses of an organisation’s Critical Data Elements. Business glossaries are essential for keeping everyone on the same page. By contextualizing and defining individual data elements, they improve business understanding, save time on searching for reports and prevent misuse. Business glossaries encourage better data-driven decision making and are an essential part of Data Governance.
Business glossaries define and contextualize critical data and reporting elements for the entire organization. They’re written in accessible, plain text and will often cross-reference terms for greater clarity. By contrast, a data dictionary is a table or set of tables serving as a centralized repository for technical metadata. Data dictionaries are rarely used outside of IT.
The first step in building a Business Glossary is to map out your critical business processes. Ensure that each data element, business term, KPI and metric used in the process is documented and defined. Include definitions of how data elements are being formatted, who is managing them, and where they are being stored.
Critical Data Elements (CDEs) are data points that directly advance your corporate strategy and concrete business goals. A good way to establish your CDEs is by mapping out your critical business processes and identifying which specific data elements are involved. You should also prioritise data elements used in critical reporting, either internally or externally for regulators.
You should develop the personnel and processes of your Data Governance program before investing in tools. Be sure to align your Data Governance policies with corporate strategy, identify concrete business objectives and clarify roles and responsibilities before you begin shopping for software. Tools should only be seen as a means of empowering your staff to achieve their goals more efficiently, not as a way to secure engagement for your Data Governance program. We advise waiting until you’re above a level 3 in the Data Maturity Assessment to begin browsing.
Understanding the different kinds of data and why they are important helps you to understand what needs to be done to drive up performance with data. The main data types to understand are:
- Master data – defining the key relationships our business has
- Transactional data – recording and reporting the events that occur
- Metadata – think of column headings; this is “data about data” and helps provide context and meaning
- Reference data – a subset of Master data, this is static or slowly changing data that helps us enforce standards
- Big data – you’ve probably noticed the volume of data increasing recently, alongside its variety and velocity. Big data describes huge data sets that can be mined for insight
- Unstructured data – pictures, text, audio and video files lack the order and structure you have in rows and columns in a spreadsheet, but there’s value locked in here that we can unlock
You’ll need to govern all different kinds of data, hence understanding what kind of data exists is important. The most important thing is not the type of data, but the value that data brings to your organisation. Many data governance initiatives are poorly scoped. All data is treated equally. This leads to a misallocation of resources. If you want to succeed you have to align your data initiatives against the main business ambitions and objectives. Identify the most critical data and govern that first.
Great question! Unfortunately, no. Many firms make the mistake of shopping for technology before they have adequately set up their data governance framework. The first thing to do is to establish the principles and policies for your business. Then, assign the right people into data governance roles, and train them to succeed. Create repeatable processes that they can follow to improve the data in your firm. Once you have that working, you’re in a position to select a great data governance tool to automate the process.
There’s no “one-size-fits-all” answer, the most important thing is to get it to fit your existing business organisational structure. At a minimum, you’ll need the following roles:
- Executive Sponsor – a C-suite leader that will push the agenda to do more with your data
- Data Owner(s) – Senior leadership that must take accountability for data. Typically they will feel the pain when data is bad quality
- Data Steward(s) – subject matter experts that are responsible for documenting and improving enterprise data
- Data Management – technical skills to measure and improve data quality, or help designing data architecture
You’ll need to change the behaviour of every single employee to succeed. These are not “data governance roles” per se, but you must drive change to data consumers and data creators across your business. Everyone’s interaction with data is going to be changed.
A data governance council sits between the Executive steering committee (providing oversight, funding and business direction for your data program), and the data governance team (rolling their sleeves up and doing the work to improve enterprise data).
The data governance council consists of data owners, driving the business direction behind improving data. They are there to break deadlock decisions that the hands-on teams can’t resolve. They’re also able to direct resources to fix data challenges. In addition to the data owners, you’ll need senior I.T. leaders on this council, who will be able to deliver changes to Enterprise Architecture and have systems expertise.
Consider the following questions:
- Is your I.T. team responsible for setting expenses policy?
- Does your I.T. team have oversight over the number of days annual leave your staff are given?
- Are the I.T. team crafting your go-to-market approach, or delivering customer service?
- Does I.T. set and execute enterprise strategy?
Most probably you answered “no”. I.T. are there to help us leverage technology to support the business goals and ambitions. Data resides on that technology, and we will need I.T. support to enforce the rules and deliver against our data expectations. However, the I.T. team are not experts in the business processes or responsible for supporting our customers and employees. As such, we must get business people to “set the rules” around our enterprise data.
The goal of data governance is actually to return value to the business. If we set up our data function in support of critical business objectives then we can demonstrate and measure a return on investment. That said, we will need to spend money up-front to change our relationship with data. Some elements of cost:
- Staff costs – we will need to invest staff time in documenting and improving data
- Consultancy costs – our research has demonstrated the benefit of engaging external experts. They don’t come cheap.
- Training costs – 61% of firms do not invest in data governance training. They are missing out and perform worse than their peers
- Technology costs – you may need to purchase tools to manage data quality or to document your metadata
It is imperative that you build a business case for data governance so you can demonstrate the work you are doing it adding value. Unless there’s a return on investment, don’t kick off a data governance program.
Still have questions? Want to learn more about Data Quality, Data Maturity, or the bigger picture of Data Governance? Not sure if your company is even ready for Data Governance in the first place? Cognopia is here to help.
With free Data Maturity assessments and a range of bespoke consulting services, Cognopia’s team of Data Governance experts will help you craft a bulletproof Data Strategy unique to your organisation’s needs. Drop us a note below and find out what Cognopia can do for you.