Turning raw data into gold: the journey to data maturity

Data has vast potential to transform business and generate customer success. At best and used wisely, it’s magical, identifying efficiencies, delivering productivity and new business opportunities. But research also shows a learning curve to generating great results – and 75% of firms are immature in using AI/ML to exploit data.

Forrester Research tells us that up to 73% of the data companies collect today is left unused for analytics. What’s in the way? At root, we’re looking at another cultural shift. While the first wave of data extraction was around collecting information, we’re learning that businesses must also become more data-centric to unleash its true benefit.

Adopting organizational learning

One report claims that just 10% of organizations benefit significantly from their AI investments. Companies that succeed with data seem willing to change company processes to facilitate organizational learning with AI. This data-centric vision requires a defined data strategy and cultural change to build data literacy. It’s about humans and data working and teaching each other.

“With organizational learning, the odds of an organization reporting significant financial benefits (from data) increase to 73%,” says Boston Consulting Group. It says that enterprises in which humans and machines learn together are five times more likely to yield substantial business benefits.

These benefits can be tangible. For example, Orange helped specialty chemicals company Evonik accelerate innovation by optimizing its data handling and data education infrastructure. It now develops new chemical formulae up to 80% faster than before while also reporting improved customer experiences.

Data-centric approach

But while some enterprises are already making good use of their data, others don’t. “Data is inherently unintelligent. It is useless if you don’t know how to use it, how to deal with it. Algorithms are where data’s true value lies,” says Peter Sondergaard, SVP, Head of Research, Gartner.

The trick is to adopt a data-centric approach. Unlike data-driven, in which data is gathered, data-centric designs businesses around the data. It builds ecosystems that derive customer and business value from the data you gather, ensuring your information is in the right place to generate such insights. Gartner also notes the need for flexible data governance, enabling organizations to respond quickly to opportunities yielded from data.

It’s important to consider how we got here. For most businesses, data harvesting happened by accident. What began as internal sales and operational numbers were soon supplemented by customer data and grew as other data stacks were added to the mix. Some of the challenges to data exploitation include the need to maintain information privacy and security, the lack of data interoperability across systems, and the limitations of using legacy technology.

Managing and structuring data

These many challenges mean much of the data that has been gathered hasn’t yet been properly structured – which means it can’t be effectively analyzed or compared. IDC predicts that by 2025, 80% of the world’s data will be unstructured. That’s a huge amount of information wasted, given that just 0.5% of these resources are analyzed today.

Throwing humans at that problem is an expensive option. Data scientists and engineers are in huge demand, so they generate significant business costs, but so much of that expensive time is dedicated to structuring, managing and cleaning the information to make it usable rather than figuring out how to turn that data into useful insights.

Machine learning (AutoML) can be used to automate some of this work, freeing up time for mature data handling. A Snowflake report notes several trends that should help companies get more from their data:

  • Consolidated data platforms and easy-to-use ML tools are evolving to bridge the gap between AI and analytics
  • Data clouds and automated data management systems help optimize even unstructured data
  • The evolution of feature stores from which data scientists can gain off-the-shelf ML solutions for swift deployment at scale
  • Rapid improvements in tools and frameworks for ML – including automated systems to optimize existing data
  • Information can also be quickly optimized using speech recognition, natural language processing or image recognition tools to create insights from raw data flow.

In other words, data scientists are building off-the-shelf tools to automate some aspects of data management. They are making these available via ML feature stores that allow enterprise data teams to select and deploy the AI components they need to exploit existing data assets. This “composable” approach sees companies combine data analytics and AI capabilities from multiple vendors rather than being locked into a single vendor.

Improving data governance with AI

AI can also play a role in data governance. Traditionally, data warehousing is centralized, but digital businesses are frequently cloud-based businesses, and the danger is that information becomes gathered in non-interoperable silos. Images, video, PDFs and an array of data in specific industry file formats sit idly in data banks yielding no insight because they remain unmined.

Orange Business is very good at making sense from these vast collections of assets. And, of course, turning such assets into consistent, machine-readable data that can work across different ML systems opens the door to deliver more value from the data you hold.

Ultimately, AI alone isn’t enough. To develop deeper insights and yield data-driven opportunities, mature organizations must focus on data-driven decision-making as a business competency. Understanding how people, machines and data work together to make decisions is an important part of this. It requires deep cultural change, flexibility, data literacy and the capacity to ensure the data you put into your systems is relevant, useful and accurate.

Find out how Orange Business can help you use ML to give meaning to your data.

Jon Evans

Jon Evans is a highly experienced technology journalist and editor. He has been writing for a living since 1994. These days you might read his daily regular Computerworld AppleHolic and opinion columns. Jon is also technology editor for men's interest magazine, Calibre Quarterly, and news editor for MacFormat magazine, which is the biggest UK Mac title. He's really interested in the impact of technology on the creative spark at the heart of the human experience. In 2010 he won an American Society of Business Publication Editors (Azbee) Award for his work at Computerworld.