The 6 pillars of data maturity

In our previous blog, we already explained that you’ll have to put in some effort if you want to grow your data maturity. Now, let’s take a closer look at which factors are important when we assess data maturity in an organisation. Because data maturity is not just an indication of how much data an organisation collects. It reflects the level at which a company makes the most out of their data, and the extent to which it uses data to process, analyse and utilise informed business decisions.

The more data mature an organisation is, the better it will be able to use data to predict opportunities, threats, and challenges. The data maturity level of an organisation is determined by six key pillars:

• The organisation
• The people who handle the data
• The data processes that are in place
• The aspects of the data product
• The technology that is used
• The data itself

Organisation

Data maturity is firstly determined by the organisation itself. An organisation always has a business strategy. But that doesn’t mean the strategy also gets translated into digital or into data and AI ambitions. The data roadmap determines the path the organisation wants to take and the goals the company has regarding the use of data. Some companies see the use of data purely as a compliance activity.

Others drive innovation with data .In those organisations, management publicly voices the importance of AI and data and the RPO of data initiatives is clear.

People

People play an important part in shaping the organisation. In traditional organisations, there are more managers that engineers. And the teams get limited autonomy to execute on the data strategy. The skill level of people is limited to traditional BI and people do things mostly “because that’s how it’s always been done around here”.

In mature data organisations you see a culture of continuous improvement and of learning. Data science techniques are being applied, teams have autonomy to deliver value end-to-end and engineers are seen as thought leaders in their field.

Process

Next, it’s pertinent to look at the data processes in the assessment of data maturity. There is usually not a lack of processes in an organisation. But do you have the right processes that actually drive business impact? In low maturity organisations, processes are mostly meant to cover asses. Ceremony is put in place so every manager can claim they’ve done their “due diligence” to make the data initiative a success. Often “Agile” is thrown in as a buzzword to virtue-signal to the world that you’re actually innovating. But often it is AINO: Agile In Name Only.

Modern data organisations apply techniques that enable teams to iterate and learn based on actual client feedback: Code is reviewed by peers, a shift-left approach to testing is followed, deploys to production happen 10x per day. Security is seen as part of the entire Software Development Lifecycle and security tooling is automated.

Product

In the end, no data organisation can stay alive long-term without shipping valuable data products to the (internal or external) customer with some predictable cadence and reliability. Some companies do have a lot of patience with their data organisations, and invest millions before expecting a first return. But sooner or later, one or more data products need to be delivered that actually add value.

In high data maturity organisations, data is leveraged for both internal and external-facing data products. And the health of those data products is always clear. SLAs are set and monitored and only a small amount of changes lead to failures. Components are made highly available and a disaster recovery plan is set.

Technology

Technology is usually the pillar of data maturity that sees the most heavy investments. This is exciting and worrying at the same time. We see quite a few organisations bringing modern cloud technologies but keeping outdated processes and outdated organisational structures without clear objectives. If you don’t know what you want, a cloud won’t get it for you either.

In organisations that have this pillar right, we do see cloud is successfully leveraged for both rapid experimentation and for industrialisation. Migration plans are in place, and for new use cases, cloud is often considered as the first option. But it’s not just about adopting cloud technologies. These organisations make rational trade-offs in build-vs-buy decisions and explore infrastructure-as-code principles. Architecture design is a balance between business needs and technical complexities and trade-offs are made between fully managed services and more lower-level cloud components.

Data

Last but not least, the actual data also influences the level of data maturity. Even with the best technologies, and the best processes, if data is still locked away in silos, or it is unable for teams to work with data of other teams, it is hard to actually create value from data. Most use cases benefit from combining different data domains into new insights.

That’s why modern data teams don’t just run their own data pipelines. They actively embrace Data Mesh ideas where other teams can build data products self-service, a data catalogue is in place to discover other data, and teams put effort into making their data available to the rest of the organisation. Strong security measures are in place to protect sensitive data and data-access is organised in a self-service way.

Where am I now?

Are you curious to find out how data mature your company is and which steps you need to take to improve your data maturity? Take our Data Maturity Scan and gain insights into how data mature your organisation is and on which categories your company scores well. Together with these scores, you will receive our Data Maturity Index that will further explain each of these categories.

Link to scan: https://dataminded.typeform.com/to/Nn0G8uPd

Shout-out to the authors of the Data Maturity Index: Jonny Daenen, Bruno Coussement and Geert Van den Broeck. Thanks for your great contributions!