AI in the business world – can it reach its true potential?

The role of artificial intelligence (AI) in enterprises continues to evolve at pace, transforming how businesses sell, how customers buy, and how suppliers deliver.

Right now, we’re seeing AI driving digital transformation across a range of industries, and nowhere more so than in the retail sector, which has faced huge challenges and opportunities as a result of the pandemic. The need to leverage big data, predictive analytics and AI has never been more important for retailers and suppliers.

Last year, analyst firm IDC predicted worldwide revenues for the AI market, including software, hardware, and services, to total $156.5 billion in 2020, an increase of 12.3% over the previous year. This is despite the economic slowdown in a year dominated by the global crisis.

At the heart of all of this transformation is data, with AI and other complementary technologies like machine learning (ML) providing new opportunities to unlock the potential of the huge quantities of data that organisations (retailers and manufacturers in particular) store today.

Data analytics, for example, is rapidly moving from human to machine-driven analysis. Until recently, developing strategies for growth involved a query-based, ad-hoc approach to test a variety of hypotheses. But this has reached its limits. With the amount of data available today, it’s simply impossible to use this methodology to identify the optimal strategy for a given product or business.

New analytics solutions allow computers to do much of the heavy lifting, powered by AI. Algorithms can analyse trillions of data points and provide prescriptive, forward-looking insights that maximise ROI.

The question though is whether AI, which promises to deliver benefits on so many levels, can reach its full potential given the possible barriers to success.

Organisational change management is the first step

Organisational change management is fundamental to success. Those companies that operate within an environment that is not prepared to embrace change are set to fail. This applies to any disruptive technologies, not just AI, but it’s clear that some are attempting to fit AI-based solutions into a traditional organisational structure, or thinking too narrowly about how to take real advantage of its capabilities.

Others are stuck in their comfort zone and rely on intuition and subjective decision making based on individual experience or previously acquired knowledge.

Some organisations distrust AI, perhaps concerned that without any human intervention they will make inaccurate decisions. For those who have based past decisions on gut instinct and don’t fully trust the technology, using techniques like running contests or ‘games’ internally to prove that AI/ML can consistently outperform human decision in the traditional way will provide a real-life test environment.

This means measuring the ROI of the decisions made, as well as measuring the ROI of the recommendations the AI/ML produced that the human did not follow or execute. Then you can use the quantified opportunity cost of ‘the road not taken’ to further support any change management and adoption.

Plugging the gap

We talk a lot about skills shortages in other sectors and AI/ML and data science is not immune from this, struggling with a talent pool that is simply not growing fast enough. Even mature AI adopters are challenged by skills gaps, according to a report from Deloitte in 2020. While the majority of mature companies, known as ‘Seasoned’, reported little or no gap between their AI needs and current abilities, 23% said they had a major or extreme need, higher than the less mature organisations.

Organisations, mature or otherwise, big or small, can take advantage of the speed, efficiency and power of AI technologies, creating a level playing field in many ways.

There’s also the opportunity to transform existing roles of information workers from preparing data and information for others, to creating algorithms and automation that produces accurate and actionable recommendations, instead of just information for further interpretation.

The challenge is in combining domain expertise and real world experience with AI/ML brute force algorithms and compute (AI + IQ). So the need for solutions that emphasise sector knowledge and grey matter capture, and automate the data science algorithm building without the need for an army of data scientists, has never been more relevant or timely.

To overcome resistance to change or to scale accordingly to take advantage of AI, businesses should think of themselves as a data democracy where information is transparent and available to all. And if they haven’t already, move towards a more collaborative organisation and away from siloed decision making. Sadly, there are many businesses that still operate in this way.

With organisations prepared to embrace change and evolve to become a data-driven environment, AI can become a powerful tool to help give them the edge and deliver on its true potential.