It’s not about Data Maturity, it’s about Metadata Adoption

Governance Is No Longer The Drive

For decades deployments of software solutions were measured mainly by speed and cost. As long as software was deployed in or around the predicted time and within or around the predicted cost, it was considered successful. Post-deployment metrics wern’t measured or scrutinized the same as the actual deployment. The huge increase in the number of software solutions a company has meant that post-deployment metrics became very important to measure success and to inform decisions around renewal or replacement of solutions. The key post-deployment metric is adoption. 
Adoption has become the golden word in enterprise software. The speed, cost, and technical success of projects remain extremely important and they often have a material impact on adoption, however, the real measurement is whether the intended users are actually using the solution in its intended way and at the expected rate.

When we consider how to measure maturity levels it helps to seek a variable that shows clear linear change from stage to stage.
When observing the maturity level of companies’ metadata management (MDM), there are 5 clear stages and adoption appears prominently as a key indicator of progress at each of them.

This is closely tied to several shifts happening in the Metadata Management space. The first shift was metadata going from not being used at all to being used as a utility of governance, it is now slowly becoming a utility of efficiency, transparency, and business growth. This is a fascinating topic that we will cover in depth another time…

A New Maturity Framework – Metadata Maturity

Let’s  look at those 5 maturity categories and explore the changes in  adoption across the maturity spectrum:

• Stage zero – Organizations with little to no awareness or utilization of metadata.
• Aware – Organizations taking the first steps in recognizing the existence and potential of metadata.
• Data-Driven – Organizations that recognize and use their metadata for technical/ IT uses only.
• Governed – Organizations that utilize metadata widely for governance and advanced use cases.
• Analytics-Driven – Organizations that have adopted MDM solutions specifically implemented to enable an analytics-driven culture.

It should be fairly easy for any organization to identify the stage it is at today and ascertain whether that is the current desired state or if additional investment in metadata utilization is needed.

As we dig deeper into each of these stages, each stage can be categorized based on several attributes  to determine its position on the maturity scale:

• Documentation – What method and solution is used to document metadata
• Utilization – What metadata is documented.
• Consumers – Which users and business functions utilize documented metadata.
• Process – The business processes applied to documenting metadata.
• Connection – How does the documentation connect to the data.
• Analytics Accessibility – How key business metrics and business questions are documented and connected to metadata.

The New Metadata Maturity Model

Here’s the approach I would use to determine the stage in which a company exists and the next steps in its progress towards an analytics-driven culture:

Stage zero:
Companies at stage zero either lack awareness of metadata entirely or see no value in it. Plenty of either very early-stage organizations or very low-tech firms  have done just fine up until now without needing to invest in this area.
These companies typically have a single source of data that is limited in scale, and the overall maturity of their data and analytics (D&A) landscape is very low. These organizations have no documentation for their metadata, and, as such, the remaining attributes are irrelevant. No utilization, no process, no connection, and zero accessibility.
The next phase after stage zero usually happens when companies start facing pains due to an increase in size or scale of operations. It starts with siloed projects to document metadata to align on metric definitions and to reduce the risk of people leaving the company and taking their analytics know-how with them.

Aware:
The benefits of documented metadata are becoming clear, but the processes are highly informal. Documentation is still done in silos and within static documents such as Google Docs, Git, or Confluence with no connection to data systems and no process around updates to documentation. The metadata that is being documented is limited to only the most used tables and columns. Moreover, the focus is only on the data that is used by technical functions such as Engineering and IT. Typically, these organizations also have pretty constrained access to analytics due to  limitations around information sharing and transparency (specifically about data and analytics).

The next phase is the beginning of formalization. As Governance and Business use cases for data start creeping up, the need for more transparent access to data increases. This is usually the stage where organizations start looking at formal solutions. And take the first steps in pushing data to be used by every function.

Data-Driven:
This is the first stage where adoption  becomes a key factor because adoption of data and analytics solutions is a proxy measurement of how data-driven a company is.
Almost every organization expresses a desire to be more data-driven but the definition of data-driven is extremely subjective and companies with widely different landscapes and processes often define themselves as data-driven.
In this case, Data-Driven is not the final stage because data is only half the story. Without instructions on how to use it correctly, its value is questionable, and it can even cause damage (I like the analogy that compares data to baking components. Without a detailed and accurate recipe, your cake will not come together the way it should or could). Companies at the Data-Driven stage have implemented, at minimum, a data dictionary. They’ve mapped all their schemas in multiple databases down to the column level and have opened access to this documentation to additional functions such as governance and finance (being only a data dictionary, it still does not really serve analytics functions). These data dictionary solutions still rely heavily on manual intervention, so updates happen on monthly or quarterly schedules using data extracts without live connections.

The next phase from here takes companies into the realm of Enterprise Information Management.  This is where governance frameworks and business processes become critical to the organization’s growth and stability, so formalization of metadata documentation is no longer a ”nice to have.” That’s the “Governed” stage…

Governed: 
Up until the recent few years, this stage was reserved for Fortune-500 companies, federal or government offices, and large, compliance-heavy conglomerates. 
They implemented enterprise data governance and data cataloging solutions because they had to, not because they wanted to. The difference might seem small, but the implementation process is drastically different, and the adoption is notoriously poor.
With full data catalog solutions, the organizations have formal processes to document data and business glossary. They embark on projects to map as much data as possible and tie the business glossary directly to data (albeit still very manually). At this stage, companies recognize the competitive advantage they can get by providing transparent, reliable, and trustworthy data to be used by analytics and business teams. These companies have optimized self-service analytics within a governed and controlled environment. 
Due to the size of these organizations’ data landscapes, the process needed to be as automated as possible and to minimize manual change management.

The next phase is a big cultural shift and, as described at the beginning, happens when adoption becomes the key factor. Beyond this point, companies need to start driving process and implement solutions to drive adoption. Adoption is the only measurement of a successful culture change.

Analytics-Driven:
Using the fun baking analogy, being analytics-driven means that all the ingredients are available at the highest quality, and everyone has access to multiple recipes that are easy to follow.
If data is made available and analytics are clearly defined, formal processes and decision-making are far more intuitive, speedy, and most importantly, reliable! 
At this stage, analytics and data science teams are no longer busy redoing work or questioning the accuracy of analytics, they are able to invest time in new exploration, experimentation, and be creative with the data they bring to drive the business’ growth.
To get to this stage, companies need modern technologies that serve the business first. It’s not about cataloging data – It’s about creating a unified, global, and living semantic layer and a single language that brings data and business terms together. 
The semantic layer needs to span every data source and include every schemas, tables, and column but, in addition, it needs to cover all the company’s analytics, formalized business questions, and official business metrics in both its business context and the query to extract the result from the data. With these processes in place, every person in every function has access to reusable analytics with evidence of accuracy and outcomes that are always aligned.

It is easy to see how this phase breaks silos, drives adoption, and creates a true analytics-driven culture.

Ultimately, It’s About Analytical Maturity

Solutions that offer active, AI-based metadata scanning mean that updates are automated, alerts are pushed, and collaboration is native. These solutions connect to all data sources, on-prem, off-prem, in a public or private cloud, and provide a single lens into the one and only version of the “truth providing”. Open access to BI and Analytics solutions becomes a no-brainer at this stage. 

As you can see from the progress of these stages, investment and formalization drive transparency and trust. These, in turn, result in adoption. 

Adoption of these solutions has a direct and material impact on the company’s culture and business health.

What stage are you at, and when are you taking the leap to the next phase?

Photo by Claudio Schwarzi on Unsplash