Triggering an enterprise data management program

"Data is like garbage. You'd better know what you are going to do with it before you collect it." - Mark Twain

Data is one of my most recent passions. I’ve started working on this topic about four years ago as part of my search for purpose in life. I’ve been working in Customer Relationship Management (CRM) and enterprise software development for well over 20 years. It was getting difficult to find new challenges. So I started researching and learning about different technology and business areas and I found two that perfectly tied up both. One of them is Enterprise Data Management (EDM).

For a while data has been considered the new blood of the enterprises. EDM is a well matured topic in some industries, especially in B2C sectors manipulating vast amounts of data. Data profiling and data quality concepts – as well as data steward and data owner functions – have been there for a long time. But with the rise of the Internet of Things (IoT) and the proliferation of digital transformation initiatives, the need of well-developed data management programs is becoming more relevant. It’s also true for the companies that haven’t traditionally relied on data in the design and execution of their business strategies.

I got the first opportunity to see one of these programs work over 10 years ago while heading another project in a bank. I got impressed by the quality of the documentation of the data concepts and data processes, but at the same time I realized the amount of work lying behind it. I doubted that other companies – in sectors less data driven – would put the same effort on similar initiatives.

Things have changed since then. For different reasons, most executives are realizing that even in businesses based on strong people relationships an effective data management program is required. Unfortunately, it’s not always easy to find a good business case and most of the programs aim for risk reduction and data protection issues driven by laws and regulations instead of adding value to their current business models.

Nevertheless, as many organizations are embracing data management practices and even when we’re not completely sure where the value will come from, it should always be the basic point for any future digital initiative. If data is not under control, the outcome depending on it can’t be under control either. There won’t be any successful artificial intelligence or machine learning project if the data behind it can’t be trusted. Even more, they can lead to wrong or disastrous results.

Ok, but where to go from here? The very first step will be to assess the current maturity level that the company has in all aspects related to data management. There are several models that can help in the analysis and provide some tools to make our life easier. All of them have pros and cons, but no matter which one we decide to use, they will help us with the evaluation. Anyway, we can always adapt them to our needs.

Most of the models provide 5 maturity levels evaluating different aspects of data management in our organizations. From my perspective, the most cost-effective way to perform the assessment is to provide a questionnaire to several areas of the organization, collect their answers, and get the average out of them.

A quite easy, simplified, and free tool to start with is provided by the Victoria State Government and publicly available: https://prov.vic.gov.au/sites/default/files/files/IM3%20Tool/IMMM_v1.7.xls

By the way, I’ve been surprised by the quality and availability of documentation of the Information Management Framework of the Victorian State Government.

After that, we will need to decide which maturity level suits best our organizations’ needs and perform a gap analysis between the current stage and the desired one. This will give us the basis for two key components of our program:

The main pain points our organization is suffering in relation to our data, to be used to build the story required to sell our data management program to the company.
The starting point to prioritize and prepare a phased implementation plan for our program.

As we can see, I’ve been talking about a program and not a project, because this is a long-term exercise. It’ll be a permanent activity embedded into the organization, with several projects tackling different aspects of the data management process. We must be sure that our organization understands it.

Five years ago, I had little experience in these topics. It took me quite a time before I felt ready to explain it to my organization. Identifying a model to follow, performing an initial assessment, discovering the main pain points, and doing a gap analysis between the current state and what I believed to be the desired state were the first tasks I dealt with. My goal was to prepare the relevant story to convince our organization about the need of an enterprise data management program and to get support to develop it.

In following posts I will share my one experience related to data and technology topics.

Some additional resources:

DMM CMMI

https://cmmiinstitute.com/data-management-maturity

DAMA DMBOK

https://www.dama.org/cpages/body-of-knowledge

EDM Council

https://edmcouncil.org/page/aboutdcamreview

IBM, Gartner and Stanford Data Governance Maturity Models

https://www.lightsondata.com/data-governance-maturity-models-gartner/

https://www.ibmbigdatahub.com/blog/big-data-governance-framework-assess-maturity