Data is one of the buzzwords of the decade. Digitalization, internet and IoT, as well as evolvements in computing power, storage, cloud and related solutions have changed what we can expect from data and what we can do with it. With that, also the different data related experiments and solutions in organizations have become reality.
Given the relatively short history of the new data buzz, I've found that organizations are in quite different places in terms of capability and maturity. Yet, whether you are wondering how to start with data in your organization or already have big data capabilities and a data team working to come up with new developments, I'd like to highlight two things without which I argue there will be no success:
- Base your data infrastructure and architecture on clear business objectives
- Make data a joint effort within the organization
Sound obvious? Good. In practice, however, I argue that many initiatives are still started too grass-level, siloed and tool focused. Especially when you are figuring out how to develop your data infrastructure and architecture, you might want to make sure it is driven by business benefits across the organization.
The problem of too narrow view
Data is an asset that needs investments in the start to put the initial structures in place, but it also means continuing investments after that. This is why you want to base your initiatives on solid business benefits.
What are the solid business benefits, then? Is it business unit's need for new better analysis capabilities or customer service department's desire to get real time operative data to adjust actions? Yes and no. From business perspective these are exactly the specific, hands-on use cases that you want to find.
From technological and data management perspective, though, you quickly get lost in different needs. When there are multiple eager departments to build up their own data capabilities, you might find your organization doing parallel overlapping projects that can lead to sub-optimization and eventually cost money, as well as build up complexity in your data architecture and management. From IT department point of view the business benefits are about cost efficiency, keeping the system complexity in control and have clear policies for governance, for example. So, how to balance between answering different needs and still keeping your data in control?
Your data solution does not need to be a trade-off.
From technological point of view, for example cloud native data platforms can provide a nice basis to start building a centralized solution that could still give flexibility and agility to answer the different business needs. In fact, our next blog will enlighten that in more detail.
Yet, what is essential to acknowledge is that the best-fit result and roadmap will still look different depending on your business objectives, maturity, legacy, data needs and operating environment. Thus, implementing new technological solutions does not automatically give you the advantage. To build a system that delivers on the business goals of your organization and yet is technically created in a way that is manageable on enterprise level requires an overall understanding of your organization's data ambitions. That is why data is not solely business nor solely IT department initiative but needs a joint effort between different departments.
Start with the expected business benefits
When starting to build a picture of what your dream data infrastructure and architecture solution might look like, you need to get a grasp of what the different data needs are in the organization now and what's the direction you want to be heading towards. Whereas the specific use cases, like the ones mentioned in previous chapter, describe concrete department needs, separately they are not showing you the full picture of expected business benefits or the possible directions in harnessing data potential.
Try involving the different departments in your organization to get an inventory of the current and overview of the potential future data utilization needs. To avoid narrowing your sight too quickly into tool-led thinking of “I want that solution”, you might want to start in a more general level, clarifying together what are your business goals and vision, and only then continue to specific data needs:
- How do we make our revenue, what is the added value we aim to provide to our stakeholders?
- What would we want to achieve and be better at there?
- How could data help get us there?
Here for example, Bernard Marr1 provides a quite nice rough categorization that you can use as food for thought and as a frame to discuss how data could help in reaching business objectives:
- Using data to improve your decision making: In other words, data can drive business decisions. This is about data guiding people's understanding and decision making in strategic, tactic and operative levels. It is about e.g. real time reports and visualizations for better situational awareness or finding new ways to combine and study data to get insights that could not be found otherwise.
- Using data to drive operational improvements: Data can drive efficiency. It can be used to optimize or even automate business processes and resource usage or drive the efficiency and transparency of your whole value network. Heineken2 sets an example of using data for targeted and customized marketing, driving marketing efficiency both to cut costs and drive uplift in sales.
- Using data as an asset itself: Data can be the business. It can be used to create new smart products and services to your stakeholders or be a monetizable asset itself. Various “My data” services are a great example, and there are also other stakeholder groups to which you could provide added value.
The advantage of putting together the different departments' views is that you'll get visibility beyond individual requests and form a bigger picture of ambitions regarding data. While it does take a step back from discussing how to create exactly a specific use case, it will benefit also the business in the long run in the form of data infrastructure and architecture that better enables supporting their needs and development. Also, as ICT systems rarely work independent from other parts of the organization and many departments may have the same data needs, driving communication can lead to heureka moments. This kind of process can also raise the level of understanding and push towards data driven culture.
Business goals drive your journey
Understanding of how you aim to benefit from data gives you a possibility to meter expected results, and it also makes it easier to communicate and argument to your stakeholders the investments needed. But, getting a larger picture of where your ambitions are gets you to the question that starts determining also what the investments actually are:
What are the current reasons for why we are not fully utilizing data in a way we'd like?
Now, this is where you can start assessing the design of your data infrastructure, architecture, management and governance practices. It gets you to the “how”, starting to raise questions like
- What kind of data do we need to be able to answer the business needs?
- Do we already have the data we need, is it good enough in quality, can we leverage external data or do we need to start collecting data?
- What would be the fitting way to collect, store and process data and how are our current capabilities answering the needs?
- How should we provide access to data?
- Is there something we are currently doing incompletely or simply not doing when we should?
This starts defining your business-driven data action roadmap and giving you requirements for data architecture.
Bernard Marr: “Data strategy -How to Profit from a World of Big Data, Analytics and the Internet of Things”. 2017, Koagan Page Limited