The Role of the Cloud in Building a Modern Day Analytics Platform

In the current environment, businesses need to move with both speed and precision. To do this, many recognize the need to build a modern data analytics platform and establish a data culture to empower data-driven decision-making across the organization.

But what does being a modern data-driven organization really mean? What are the hurdles to becoming data-driven? And what role does the cloud play in helping build a modern data analytics platform?

For answers, we turned to members of the IDG Influencer Network, a community of industry analysts, IT professionals, and journalists. Although their viewpoints varied, as would be expected, there was surprising unanimity in their responses to these questions.

“A true data-driven organization is built around the core values of protecting data as a real asset,” said Scott Schober (@ScottBVS), president and CEO of Berkeley Varitronics Systems Inc. “It is important to use emerging technologies such as analytical forecasting, machine learning, and artificial intelligence [AI] to separate the valuable needles from the digital data haystacks.”

“The modern data-driven organization requires guiding people to new ways of working with data on a daily basis by asking questions and self-serving analytical insights,” noted Isaac Sacolick (@nyike), president of StarCIO. “Wrangling data in spreadsheets should be replaced by robust data operations, enabled by data prep tools, and improved quality through proactive data governance.”

Sacolick’s viewpoint was seconded by Will Kelly (@willkelly), a content and product marketing manager: “Being a modern data-driven organization is about making corporate data accessible to your business users via self-service tools anytime, anywhere,” he said. “It means putting in the tools and platforms for data management, governance, and security on the back end. User experience is a paramount requirement to open your data up to business users.”

For Steve Prentice (@cloudtweaksteve), a technology integration specialist, “being data-driven is the yin-yang to intuition and gut feeling. Neither by itself is sufficient to base a company’s future on, nor should one displace the other. A great deal of business success comes from having the right data at the right time but also knowing how to interpret it and what to do next.”

John Moore (@ACollaborator), the founder of Trust Enablement, agreed: “I often refer to being data-aware vs. data-driven, but both are far more important than being data-blind,” he said. “Being data-aware means you leverage a mix of quantitative and qualitative to track your progress towards your business goals.”

It’s a difficult proposition, Moore freely admits. His advice? “Implement data hygiene, centrally manage the data, create processes to maintain it, educate your team on why it matters and how to do it, and use the data to guide your way,” he said.

In the past, data-driven usually meant having transactional data from traditional business systems and doing analysis on those transactions to intuit trends and make business decisions, observed technology evangelist Ed Featherston (@efeatherston).

“But in today’s world, a truly data-driven organization is thinking about what data is needed to drive business decisions and innovations,” he continued. “With that, applications and technology are planned and designed around those business requirements to ensure that the right data is collected and access is given to the right individuals to be able to perform analysis and make decisions quickly.”

Cultural change tops list of hurdles

When it comes to the hurdles that organizations confront in becoming data-driven, there was pretty much across-the-board agreement about what leads the list. Featherston spoke for many of the influencers when he said the “biggest and hardest” hurdle is cultural change.

“Becoming a data-driven organization from a traditional IT structure is a very disruptive process with lots of change,” he continued. “Ultimately, technology is easy; culture is hard. This is quickly followed by skill sets in the field of data to help understand what data is needed that provides value, where it needs to go, and who needs to see it. Finally, avoiding falling into the trap of building a data swamp, where you just start collecting data from everywhere, put it in a data lake, and hope and pray analysis of the data will provide value.”

“Not unlike building a culture of cybersecurity, building a data culture within an enterprise can be even more challenging,” agreed Frank Cutitta (@fcutitta), CEO and founder of HealthTech Decisions Lab.

“The reasons are threefold,” he explained. “First, the utility of the data is sometimes questionable. For example, the data might simply confirm what is already known, causing the ‘ho-hum’ effect from the data users. The second is that — for many reasons, some generational — there is a skepticism about using data over intuition. For example, this can be the case with mature doctors as it relates to their using AI. Finally, great pains must be taken to go beyond circulating raw data, or even data visualization, and building a story that can be embraced and socialized by the data constituents. This storytelling and training on the use of the data findings is absolutely critical if there is an expectation of empirical value of the data collection and analysis effort.”

“The challenge is a cultural one,” said Prentice. “When a change is designed and launched based on data, that change will still come up against human nature, which will often put up resistance as a first resort. . . . Management should embrace the idea of becoming data-driven in the same way it seeks to become agile and lean. These can all be buzzwords unless they are respected and understood. They cannot be expected to be a quick fix.”

Of course, cultural change isn’t the only obstacle.

“The biggest challenge most organizations face in becoming data-driven is in finding and consolidating all of their data assets into one combined entity that can then be analyzed to provide insights,” said Jack Gold (@jckgld), president and principal analyst at J. Gold Associates. “Partial data sets, invalid data, formatting errors, incompatibilities, etc., all make it impossible for companies to achieve the optimum value of being totally data-driven. It also negatively affects the ability to use analytics and/or AI to provide insights that are accurate.”

“Too many organizations either get overwhelmed with ‘big data’ and fail to utilize it effectively, or they are so focused on their core product that they only think about data as an afterthought,” added Schober. “Both approaches are a failure to capture the true value of customer data.”