Traditionally an underrated business asset, data is increasingly becoming vital to bottom-line success. From NFL teams and airlines to fast-moving consumer goods firms, leading organizations are embracing the potential of data and are using it to transform how they operate.
As many countries begin to chart the journey into a post-pandemic world, data is being looked to in order to help understand the drastically different business and social environments that companies find themselves in.
However, despite the progress many companies have made towards extracting value from data, much more needs to be done. A Deloitte survey of executives in the US discovered that a 63 per cent majority do not think their organizations are analytics-driven, with a further 67 per cent saying they are not comfortable accessing or using data from their tools and resources.
In the 15 years since British mathematician and marketer Clive Humby exclaimed that data is the new oil, the data economy has only grown. However, Humby’s warning that if data is unrefined it cannot really be used illustrates the challenges of extracting actionable insights from huge amounts of disorganized data.
“The key here is the word ‘actionable’,” explains Paul Alexander, Group CEO of data science and analytics firm Beyond: Putting Data to Work. “Data is amazing and can help businesses discover loads of really interesting things. But unless it makes a tangible difference to the bottom line then it’s pointless and costly.”
With a staggering 2.5 quintillion bytes of data being created every day, organizations can face an uphill battle to focus on high-quality data and disregard unusable information, especially for firms with legacy systems and siloed data. Searching through the countless data a company holds on its customers, employees and operations is no small task.
“Unless it makes a tangible difference to the bottom line then it’s pointless and costly.”
- Paul Alexander
Alexander and his team recently partnered with a staff resource organization that wanted to unlock the value of its data to increase efficiency by managing the work around the workforce. To achieve this goal, Beyond set out to build a unified data source that creates one source of truth and increases confidence in the data, as well as establishing a single reporting platform.
“Once we reconciled the data, we were able to create supply and demand forecasting models that for the first time enabled the organization to see future trends and make planning decisions ahead of time which resulted in optimized performance,” he says. According to Alexander, the additional revenue of this approach is estimated to be more than US$198.9 million.
Jason Jercinovic, Vice President of Technology at international change and transformation consultancy North Highland, agrees that data plays a key role in improving operations and uncovering new opportunities, but explains that business leaders need to ensure that customer, workforce and operational data is centrally located and integrated across systems.
One way to ensure the right groundwork is set up before embarking on a major data project is to adopt an insights framework. “Using this framework, start by understanding strategic priorities; brainstorming business decisions to support those priorities; identifying the business insights that will inform those decisions; defining the KPIs relevant to those insights; defining the dimensions by which the KPIs can be sliced or filtered; and designing the portfolio of dashboards,” Jercinovic says.
The value of business data is clear, with an Experian report finding that 85 per cent of organizations see data as one of their most valuable assets. However, like any asset, when it is not used properly issues can arise. The nature of many advanced artificial intelligence (AI) and machine-learning solutions means that often the unconscious biases of humans are replicated when data is analyzed.
“There are literally thousands of horror stories when it comes to data,” Alexander says. “The fact is that AI algorithms learn from given, often external, sources of data, and therefore their actions will naturally reflect the leanings or affinities of the information these sources contain.”
“The fact is that AI algorithms learn from given, often external, sources of data, and therefore their actions will naturally reflect the leanings or affinities of the information these sources contain.”
- Paul Alexander
Bypassing algorithmic bias can be a difficult task to accomplish, with even Amazon and Google failing at this hurdle. Tech giant Amazon was developing an AI recruiting tool that was trained to assess applications based on resume patterns over a 10-year period. Yet as men account for the majority of applications and staff, the AI trained itself to assess male candidates as being preferable.
Working with a talented business partner who has expertise in AI and machine learning can go a long way to protecting against potential pitfalls, as can checking progress towards preset benchmarks.
With new data being collected on a daily basis, companies need to stay ahead of the innovation curve and ensure this data is being put to the best use. For organizations that fail to put a priority on establishing a comprehensive data strategy, it’s unlikely they will succeed in an era where data is an essential asset.