This blog was authored by: Ben Jarvis, Data and AI CTO | 3rd February 2025

 

Following our first article, “Beyond the Hype: The Real Value of AI,” where Martyn shared insights on cutting through the noise, this second instalment focuses on a crucial aspect of AI implementation that I always discuss with customers; building a robust data culture within your organisation. 

 

AI is a hot topic; however, the true foundation of successful AI integration lies in establishing a strong data culture within the business. In this article I will cover the essential components and steps required to build a data-driven culture that empowers businesses to fully utilise the potential of their data. 

Key Components of Data Culture

Data isn’t just a technology issue; it requires fostering a data culture throughout the organisation and equipping teams with the right tools for decision-making. We focus on four key areas when helping organisations develop a data culture:​

Key Components of Data Culture 

People are the single most important part of our transformation into a data-driven organisation. We must secure executive sponsorship to ensure the success of our programme and consider strategies to guide the business towards becoming data-driven. This includes planning for change management and communications, as well as establishing a learning and development pathway to train teams on the new technology being implemented.

As data usage expands throughout the organisation via various tools and processes, it is imperative to implement effective governance and security measures to prevent potential data breaches. We must classify all sensitive information appropriately and establish processes to monitor and secure the usage of this data. While governance is crucial, it is essential to position it as an enabler rather than an impediment for teams. 

It is essential that the business can find the right data at the right time, to enable this we need to ensure that we have an effective data catalog / marketplace in place to ensure that users can identify what data is available, understand how that data can be used, where it can be accessed and identify who the key point of contact is to support the usage and evolution of that dataset. It’s also essential that we use a common taxonomy across the organisation that is familiar to all members of the business to ensure that users can receive the right answers when searching for datasets. 

As a technology organisation it feels unnatural to leave technology until last however, all of the above elements are key to ensuring we can implement an effective data culture and technology is there merely as an enabler for those capabilities. A scalable and secure data & AI platform that contains the tools for developers, data scientists, analysts and business users to leverage data in an effective and efficient way is key to making the process work. We also need to ensure that the platform is future-proof and has the capabilities available to enable the adoption of technologies such as Generative AI, as the business needs require it.

Maturity Stages of Data Culture

Implementing a data culture is a gradual process that involves technology changes and adjustments to the culture and processes within an organisation. Different organisations may encounter various challenges, particularly related to the human aspects of the transformation. The following is the journey we typically see customers move through to reach the desired level of maturity, it’s important to note that this is a generalised set of stages and each organisation is different and may begin at a different stage of the journey: 

Maturity Stages of Data Culture

Typically organisations start with a community of analysts/savvy business users working in tools like Excel and Power BI to dabble in analytics and utilise direct connections to source systems to retrieve the data needed. In most cases this stage results in a situation where some great analytics are being delivered but there is no control or governance in place meaning multiple sources of truth result and different areas of the business come up with different definitions for key metrics, resulting in distrust when attempting to make decisions using the data.

The first stage of gaining control of data within the organisation is to stand up a centralised data team that is able to identify key use cases/metrics and implement them in a governed platform that incorporates good software development practices. The centralised team will generally deliver a set of “corporate reports” that contain key metrics, with the necessary trust being established to ensure that the data is correct. This data can be used by the management team to make clear decisions on trusted data.

Once a centralised capability has been established and the organisation has the necessary trust in key metrics/datasets we can begin to open the data platform up to “citizen developer” / “power” users that are keen to use the data within the platform to build their own use cases to share within their own teams, whilst ensuring the right guardrails are in place to determine whether their outputs have been through an assurance process before being shared with other areas of the business. The “power” users are generally distributed across different business functions and can aid in building up momentum within the organisation to show the true value of data. 

As the group of “power” users mature we can begin to roll self-service out further across the organisation and build communities in each business area that can collaborate and identify new and inventive ways to utilise data for improved decision making whilst managing risk and ensuring the data can be trusted. At this stage, some organisations consider adopting the data mesh approach of distributing the engineering / ownership of data within individual business areas however, the merits of that approach are a topic to be discussed in a separate blog. 

Target State

The result of moving through each of the above stages is a platform that is able provide the business with the data and tools to use data in a way that allows them to take the right actions at the right time, using data that can be trusted. These actions can be driven by conventional analytical reporting, machine learning models, generative AI models or any other technology – what’s key is we have a trusted foundation of data to make those decisions on. 

Customer Journey Examples

Every organisation should embark on a data culture journey, but it’s important to recognise that we are all at different stages of maturity. This journey isn’t necessarily linear and can be changed by factors such as acquisitions and business modernisation. The focus of the business—whether on data science and AI or data analytics—and the available resources also play a significant role. 

 

Here are some examples that highlight the diverse approaches and progress in developing a data culture:

Our Goal 

Ultimately, our goal is to help organisations develop a platform fully aligned with their unique business objectives by understanding key questions and identifying actions that influence success. By building data products—whether analytical tools, machine learning models, or AI applications—we provide valuable insights and predictive capabilities.  

 

A strong data culture supports not only custom AI use cases but enables the use of Generative AI tools like Copilot and Genie. By building high-quality data with the right metadata, organisations can have the right base to put AI tools over the top and provide accurate results.  

 

Implementing a data culture requires a strategic approach, starting with understanding the potential of data, laying a strong foundation, and gradually advancing to more sophisticated uses of AI. We guide organisations through this journey, ensuring they achieve their business objectives with data-driven insights. 

 

For more information on how to build a data culture and leverage AI, either message me directly or fill out the form below.