Spending on big data and analytics products was expected to surpass $200-million last year, according to the International Data Corporation (IDC). By the end of 2023, the big data analytics market will be worth more than $100-billion, and in a Bloomberg Businessweek Research services survey, 97% of respondents claimed their businesses have invested in big data analytics. Big data, it turns out, is very big indeed and set to get even bigger.
In the midst of a global epidemic, interest and investment in data analytics have only grown. Companies can turn insights into profits, and visibility into cost-saving measures. Whole businesses exist because they’re able to convert data into action, or help others without the means to themselves, collect and collate enough variables to glean useful information from them.
Leveraging data, whether for modeling the potential outcomes of decisions, informing real-time decision-making, evaluating and assessing risk, or streamline processes and unlocking efficiencies, is a global trend rapidly gaining momentum. Yet, less than 30% of organisations self-described as having a “data culture”, and in 2019, only 31% of respondents to a survey by New Vantage Partners called their firms “data-driven”, a decrease from the year before. In 2020, only 23% felt comfortable with data analytics.
The BearingPoint CDO survey
Three-quarters of companies find adopting big data and artificial intelligence (AI) analytics a serious challenge, according to Gartner. Why? The problem is not a lack of technology. Instead, it’s by and large cultural. In 95% of cases, executives say it is people and processes that stand in the way. From the challenge of fostering universal data literacy (something 65% of respondents cited), to skills deficits (49%), and even simply resistance to change (56%), executives looking to embrace data analytics face an uphill battle, with many of the stumbling blocks coming from within.
When securing and retaining staff with the requisite skills, being able to secure resources to drive analytics and data literacy programmes is critical. Without it, business units that could benefit from data-driven insights can find themselves siloed from the data required for meaningful results (whether deliberately or inadvertently) and can struggle to convince their organisations investments in data analytics and AI — which may take time to show quantifiable results — are worth the resources and patience they demand.
There’s a tendency for businesses in certain sectors to want to employ AI even before they’ve worked out which problems they are hoping to solve. Basic AI literacy in all parts of organisations matters because it can help make it clearer whether AI is even the right solution to consider. But it does more than that: it creates a core level of understanding and the shared vocabulary required to engage about data and AI, and assess its appropriateness for mining insights.
A workforce more attuned to AI’s use cases (and limitations) is also more likely to view AI as a potential life improver, rather than an existential threat. And it’s likely to help maintain realistic expectations of what AI it is actually capable of; which problems it excels at solving; how to deploy it. When AI is a tool among many, it’s enabling, not threatening. With greater comfort, come more innovative results.
Being data-driven is hard
A 2019 survey of executives as large firms found that a full 75% of employees say they’re uncomfortable working with data. A third claimed to have taken sick leave due to data-related headaches or stresses. No wonder data science initiatives fail, data-driven cultures fail to emerge, and organisational resistance remains high.
Why is it so hard for businesses to talk about AI and analytics, let alone take advantage of them? And what can be done to change that? To get data science teams to generate a return on the investment, organisations need to create human resources “bridges” that allow the data team’s abilities to cross over into other branches of the company.
Building those bridges happens through sustained and cumulative data literacy development initiatives. Not doing so leaves the various divisions of your business effectively speaking languages themselves no one else in the organisation does — like a group of people where no two people share a lingua franca. Unable to communicate their unique obstacles, let alone the ability to plot a course around them.
“The prevalence of data and analytics capabilities, including artificial intelligence, requires creators and consumers to ‘speak data’ as a common language,” explains Valerie Logan, a senior director analyst at Gartner. “Data and analytics leaders must champion workforce data literacy as an enabler of digital business and treat information as a second language.”
Most roles within financial organisations will soon find the ability to “speak data” cogently and concisely will go from a nice-to-have to an essential skill. Employees, no matter their department or position will need to have at least a fundamental understanding of machine learning, natural language processing, robotics, and deep learning.
Further, they’ll have to be able to read, write and communicate about data in context, understand sources and constructs, be able to deploy analytical methods and techniques, and talk about outputs and outcomes.
The essential skills
To be data literate, an employee needs to understand data concepts and their application; be comfortable with collection and access; assess data relevancy, and understand how relevant data is managed and synthesised. Employees must be able to do data-driven inquiry, using data tools, and to be able to communicate not just their goals, but their approaches, experiences and results.
AI has already come from providing voice-based assistance on smartphones and home speakers and playing human games such as Jeopardy or Go to guiding driverless cars, providing better-than-human medical diagnostics, and both creating and detecting deepfakes, some of which are sufficiently convincing they fool humans every time. And all of this has happened in a little over a decade.
Often the most boring or time-consuming parts of a job are the ones ripest for automation. Enhancing pan-organisational AI literacy can help employees recognise these areas and communicate them with the requisite precision that they can be automated out of their hands. This is the promise of automation aided by an intelligence the artificiality of which makes it perfectly suited for the mundane or monotonous.
The future of work will likely be filled with human-machine partnerships. Rather than replacing humans, robots and AI will remove the rote and repetitive, freeing humans up to work on other projects to which they’re better suited, those that require critical thinking, and which people tend to find more satisfying.
AI and data analytics, correctly designed and deployed will replace tasks, not jobs, and create new jobs — some of which haven’t existed before — in the process. But to get there, we need to be able to talk to one another in a shared language around AI and data. As the world’s datasets continue to grow and diversify, our ability to engage with and take advantage of them needs to increase and adapt in step.