Data-Driven Business Value
Over the last years, being data-driven has gone from being one of the big technology trends, and an actual differentiator for those living up to it, to become almost taken for granted. The industry has moved on to rave about, or fret over, how machine learning and Artificial Intelligence(AI) will revolutionize the world, potentially bringing on the end of all of humanity as an unwanted side effect.
Whether we should treat AI as mostly a threat or an opportunity, and what actions we ought to take to prevent us from shooting ourselves in the foot (or worse) with it, deserves a blog post of its own. For now, we know that AI, specifically generative AI, is here to stay. With that, the way we interact with data and information has fundamentally changed. The positive side of this is that it has greatly increased our efficiency in doing many tasks, and it is constantly unlocking new innovations and business opportunities for those who are creative and ready to use it in their own context.
You can get efficiency wins simply by adopting pre-trained Large Language Models (LLMs) such as ChatGPT, Gemini or Copilot, but to really build on top of the AI and machine learning paradigm you need to be in control of your own data. For that, you need a clear data strategy and architecture. Unfortunately, even though data-driven may have lost its buzzword status, it doesn't mean it’s a solved problem; far from it.
Handling data within a large organization, with good quality and with information security in place, is inherently tricky. Before jumping on applied machine learning and AI solutions, you need to invest in the foundation, including all the necessary capabilities such as capturing, transforming, storing, accessing and visualizing data, keeping it secure and in line with an ever-growing number of regulatory requirements. You need to figure out what processes and governing principles you need as support, to empower your entire organization, and possibly also external users, with the information they need, at the right time.
At Trice, we’re involved in a number of different data related transformation programs, where we help our customers define their data strategies and implement modern data architectures. In some cases the goal is a multipurpose, generic and enterprise-wide data platform, in other cases the need is more specific, such as gathering and visualizing actionable cyber security insights.
We came together to share our experiences of what the successful transformations we’ve seen has in common and where the biggest challenges reside. In this first blog post on the topic we’ll cover how to set goals that are aligned with business value, assessing your current level of maturity and the most common pitfalls and how to avoid them.
Start with Business Value
As with all changes, you need to know what your end goal is, before deciding on a strategy that will take you there. A well-built data platform, with lots of technical capabilities, shouldn't be the end goal in itself; it should be a tool to drive business outcomes and value. This may sound obvious, but we’ve seen many data transformation programs fail because they were driven from a purely technical perspective. Understanding the existing technical landscape, as well as what is possible to achieve in a modern, often cloud-based, environment, is one part of the puzzle. However, without a clear understanding of what your business needs are, you might end up with a platform that is not fit for purpose, or that won’t be used. In those cases, the initiative is often shut down before it delivers any value, as it is mostly (seen as) a cost.
Aligning your data strategy with concrete and prioritized business goals, such as improving customer experience, optimizing operations, or finding new revenue streams, increases the chance of broad buy-in and adoption.
The goals should be specific to your organization. However, there are some common use-cases and goals that we have seen across different organizations:
Enabling data-driven business decisions & reporting:
Empower data analysis: A modern data platform should make it easy for users to access, explore, and analyze data, without compromising information security. With integrated BI and analytics tools data can be turned into insights that inform business decisions.
Improve efficiency: Data can be used to identify inefficiencies in processes, optimize resource allocation, and streamline workflows. This can lead to cost savings and improved operational performance.
Support business reporting needs: Data needed for regulatory compliance and reporting should be easy to access. A flexible and easily extendable data platform can allow you to access the data you need, and quickly adapt when regulations change, or new requirements emerge.
Use data to drive innovation of new digital services and products:
Data-driven product innovation: Capture customer and user data and use it to validate hypotheses and guide product development, marketing strategies, and overall business development. Having data easily accessible allows you to understand your users better, which may lead to entirely new innovations and value streams.
Use AI and applied machine learning on existing data: By making data available to R&D and product development teams across the organization you can empower them to build entirely new products, for example by using machine learning to create recommendations based on a user's previous patterns. There are lots of untapped opportunities in many industries, as applied machine learning and AI make entirely new innovations possible.
You might want to cater to some of these, or all of them. But we recommend prioritizing the business goals and use-cases, to make sure you start where it adds the most value.
Assess your Current State
After setting goals, you need to assess the current state. There are many models and frameworks aiming to define and assess the data maturity level of an organization and its people. An easy and straightforward way to do this is using these three levels of maturity: data aware, data literate and data-driven:
Data aware: People and teams who are data aware recognize the importance of data, they have started collecting some, but they might not be able to make use of it, at least not to the extent of making decisions or in fulfilling business needs.
Data literate: An organization (or individuals/teams within one) that is data literate has moved beyond just collecting the data, they also have the ability to access, understand and analyze data.
Data-driven: This is the most advanced level of data maturity where data is seen as a strategic asset. At this stage you have a well-defined data strategy, technical capabilities and infrastructure that allows you to work with data, and a culture where everyone across the organization is able to use data effectively in their decision-making.
There is almost always different maturity levels in different parts of an organization. Unless you’ve already reached data-driven maturity in all parts, then this blog post might not be for you.
Defining your Data Strategy
Once you have your goals and current state understood, you’re ready to set a strategy and start the transformation, moving different parts of the organization, depending on where they stand, from data aware via data literate all the way to data-driven.
Depending on your specific goals and the business and technical context you’re in, the strategy you set for your organization will necessarily be different. Make sure your data strategy covers at least the following aspects:
Organizational setup and data ownership: Who in your organization should own the data and the data quality? Will you aim for centralized ownership, de-centralized ownership or implement something along the lines of a data mesh, with de-centralized ownership but with federated governance and a central self-service platform?
Governing principles: How will you ensure good data quality and information security? And how will you handle compliance with relevant regulatory requirements and standardizations, such as GDPR, ISO 27001, NIS/NIS2 to mention just a few.
Technology stack: Involves choosing how you implement the necessary technical capabilities and infrastructure needed to support your data strategy. For example, will you choose a cloud-based or on-prem solution, or a hybrid?
We’ll cover each of these topics more in depth in our next blog post on this topic.
The Road to Success
Data strategy transformations can fall short due to a number of reasons, very often stemming from a disconnect between business and technology and a lack of clarity of goals. Here are the most common pitfalls that we’ve seen – and how to avoid them:
Lack of clear goals and direction: Without clear goals aligned with business needs, it's hard to measure success or justify the investment. Our most important advice is to always make sure your data strategy is well aligned with prioritized business goals, with stakeholder buy-in both on the business and tech side of your organization.
Lack of adoption: Focus on initial implementations that deliver quick wins and demonstrate the value of the platform to stakeholders throughout the organization. This builds momentum and encourages wider adoption. Start simple and scale gradually, delivering value in each step, but be careful not to cut corners in building up the platform and technical capabilities, as this will become costly later on.
Poor data quality: Avoid "garbage in, garbage out" by prioritizing data quality from the start. Establish data governance policies and standards to ensure clean, well-documented and reliable data. Train the teams producing data to treat their data as a product, and make sure ownership over the data is clear.
Becoming data-driven might seem like a big undertaking, involving both investing in new technical capabilities and an organizational and cultural shift. Successfully implementing a data strategy that really delivers value will, in most cases, involve some quite fundamental changes throughout an organization. However, if done the right way, it is very possible to deliver business value early on.
Above all, succeeding with your data strategy transformation (becoming truly data-driven) will unlock new business and technical possibilities and allow your organization to end up on top of the next big shift, the AI and machine learning paradigm.
