Whether you’ve already achieved top-down, C-level support for your first data project, or you’re trying to achieve the business value that you need to do so from the bottom up: you have to start somewhere. The possibilities often seem so large scale that jumping in can be daunting. Take a deep breath – and ...
Step 1: Familiarize yourself with the data you have
Getting your data organized today will prepare your business to tackle the surprising twists and turns that the future is sure to bring. But that doesn’t mean you can start your data journey off by setting up a data platform without first considering what you want to accomplish with it.
“We’ve worked with companies that are collecting huge volumes of data and dumping it all into a data lake – but many of them have no idea what data points they are gathering,” says Kristel Demotte, VP of Data solutions at Cegeka. “They implemented that data lake with no particular goal or roadmap in mind. If you only start thinking about what you want to do with your data while you’re accumulating it in some dark corner of your IT infrastructure, you’re already too late.”
Before you can start getting actionable insights from it, you have to explore the data you have to work with. What tools and processes generate it? Where is it stored? How is it managed? What is its quality?
"It’s important to look at things from a business perspective. Explore which business questions are most relevant to your company, and then consider which data will help you answer them."
Kristel Demotte, Global VP of Data Solutions at Cegeka
“Also important here is to look at things from not just a technical viewpoint, but from a business perspective,” adds Kristel Demotte Global VP of Data Solutions at Cegeka. “Explore which business questions are most relevant to your company, and then consider which data will help you answer them.”
Step 2: Start where things are going wrong
Bottlenecks in processes, workflows that take too much time, repetitive manual tasks, places where errors tend to happen over and over again: these are all great starting points for your first data project.
Do you already have a business domain in mind? Set up an inspirational brainstorming session with key stakeholders that are close to the processes in question and invite them to toss out potential business questions to answer.
“It’s only once you have identified the right use case that you can move into the proof of value phase,” Kristel continues. “If your enterprise has never worked with data before, it’s best to start off with low-hanging fruits, such as establishing a BI environment that reports on your core business applications. Once you have that in place, you can expand it with more advanced use cases later.”
Step 3: Invest in a data future-proof data platform
Even if your company already has a data platform in place, it won’t necessarily offer the capabilities necessary to respond to your business questions today.
“Traditional data warehouses are built to store traditional data. They simply aren’t equipped to handle sensitive data, external data, unstructured data and streaming of data that is so useful for modern use cases. Even more, legacy data warehouses and platforms are often completely standalone systems. If you are ambitious, experienced with data and want to explore AI as your next innovation step, you will absolutely require a future-proof data platform that is secure and scalable.”
A future-ready data platform foresees secure data storage, is capable of scaling up and scaling out, serves a multipurpose workload and integrates with your existing IT landscape.
"The public cloud is the perfect environment for a future-proof data platform. You can begin with a scaled-down, low-cost solution and then extend it as needed."
Kristel Demotte, Global VP of Data Solutions at Cegeka
“The public cloud is perfectly suited to this type of platform,” Kristel asserts. “It’s an agile environment and compatible with a ‘start small, think big’ mindset, because even if you have nothing at all, you can begin with a scaled-down, low-cost data platform in the cloud and then extend it as needed.”
Step 4: Build your data competency center
If you seek to become a data company, you will necessarily give data a big role in your business strategy – which requires a competence center. But where do you start?
“A proof of concept is an important catalyst for your data competency center, because it drives ambassadorship at both the management level and from the bottom up. By proving the value of data insights, your enterprise justifies the budget needed for innovative data projects and develops on-the-ground data experience. If you have the right mindset, you’ll notice that several things are happening almost simultaneously: visibility, ownership, buy-in, data skills development and the use of data in business decision making.”
In order to work with data in the public cloud, introducing the agile methodology to your team is an absolute must. Waterfall development methods are less suited to data projects, which are apt to change and evolve as they are implemented.
Step 5: Make data an integral element of your broader IT ecosystem
A clear theme emerges as enterprises become more data mature: integration. “In the ideal data company, every single part of its IT ecosystem is connected – from CRM to ERP to digital workplace and beyond,” Kristel says.
“Connecting these elements is the next step after your first projects bear fruit. This is where an experienced data partner comes in handy, particularly one with expertise in business line integration, infrastructure support and managed services in the public cloud. If you’re already using data to drive processes, a partner who thinks alongside your business in innovation domains like connectivity and custom-crafted AI solutions will ensure smooth, rapid business value.”