Why predictive analytics could be a crystal ball for businesses

Predictive analytics is becoming a valuable tool for companies seeking to model the likely outcomes of key decisions before committing themselves. But implementing the tech is easier said than done
Data Intell P8 9 Illo Online 01

With rampant inflation increasing their costs and recession fears dampening demand for their goods and services, firms nationwide are busy cutting unnecessary spending and seeking new growth opportunities. 

Determining where, when and how much to spend is always important, of course, but it’s crucial in a downturn, when such choices can have a huge impact on a business’s medium-term growth prospects once the economy recovers. That’s why, in search of ways to refine their investment decisions, firms are increasingly using predictive analytics to help them weigh up all the potential opportunities and threats. 

“Risk management is never far from discussions among CFOs and regulatory teams,” says James Petter, vice-president and general manager, international, at Pure Storage, a data storage firm which helps organisations grow through digital transformation. “But, in the economic climate of 2023, every senior leader in every organisation will have risk management front of mind. They’ll be making a deep assessment of the economics within companies, their financial structures and technologies. 

He continues: “Every company has a wealth of data and most are trying to do something with it. But often they’re focused on understanding the current market conditions and reacting to these. I believe that there’ll be more of a push to look ahead as part of the overall focus on risk management. Predictive analytics will play a big part in this.”

The rise of predictive analytics comes as no surprise to Shankar Balakrishnan, vice-president for northern Europe at software developer Anaplan. He likens firms that rely on historical data alone to navigate in such tough conditions to a driver steering their car according to what they can see in its rear-view mirror. Instead, Balakrishnan argues, they need access to more data sources to model potential future outcomes and so react more smartly to disruptive events. 

Anaplan recently worked with the South Central Ambulance Service Foundation NHS Trust (SCAS), which covers Berkshire, Buckinghamshire, Hampshire and Oxfordshire, to help it develop a predictive capability. Applying machine learning and predictive insights to existing data, Anaplan was able to forecast the number of emergency calls the SCAS ambulance teams would receive at any given point. This has enabled the trust to deploy its resources more efficiently. 

The implementation headache

But what’s the best way to implement this powerful AI-based technology? “For finance chiefs, the challenge is to understand where to focus,” says Simon Edwards, CFO at software developer ServiceMax.

One good place to start, he suggests, might be automating functions in the back office. Using tech such as robotic process automation and AI-enabled data analytics not only helps to improve routine processes, cover skills gaps and increase efficiencies; it can also provide intelligence that can be fed into forecasting and planning. What’s more, this kind of automation will also enable staff to focus on more value-adding tasks, he suggests.

Implementing predictive analytics isn’t a case of ‘run once and forget’. It will take time and effort to analyse the findings, understand them and then tweak the program accordingly

Given that the commercial environment is awash in risk and uncertainty, few leaders will want to trust important resourcing and investment decisions to gut instinct. Indeed, risk management may be the top priority in times of crisis, but what if business leaders could avoid the crisis in the first place?

Whether you’re facing a pandemic, a natural disaster or a ransomware attack, making effective choices under pressure requires accurate and timely data-driven insights, notes Alan Jacobson, chief analytics officer at data science company Alteryx. 

“Successful risk management requires data as the course corrector, giving you the ability to model different scenarios,” he says. 

Jacobson points to travel and tourism as industries that are banking on predictive analytics to help them recover fully from the hugely damaging disruption they have suffered in recent years as a result of the Covid crisis. Aircraft manufacturers are using the technology to determine the most effective times to perform various maintenance tasks. Airlines are using similar systems to predict demand for particular flights and plan their staffing and fuelling requirements accordingly to improve operational efficiency and minimise disruption.

“Quality data and predictive analytics are also integral to risk mitigation across the financial services industry,” he adds. “They are invaluable for fraud detection, audit investigations and other types of advanced work.” 

Accuracy matters

Of course, the success of such efforts hinges on the standard of the data fed into the system. Insights based on faulty or incomplete inputs could mislead decision-makers and potentially cause significant harm to a business.

“Implementing predictive analytics isn’t a case of ‘run once and forget’. It will take time and effort to analyse the findings, understand them and then tweak the program accordingly,” Petter stresses. “The risk would be implementing a big program that doesn’t deliver the insights needed to help the organisation. It’s important to have clear goals when implementing, adjust as needed and constantly refocus to ensure that the business is getting what it needs.”

Indeed, numerous data problems can lurk beneath the surface, especially if users are inexperienced in handling the outputs generated.

“Accuracy and compatibility are paramount when it comes to measuring performance across different departments,” Edwards says. “It’s a common problem that needs addressing before it does material harm to a business.”

Balakrishnan agrees. “If leaders are working with inaccurate data, they risk making inaccurate decisions,” he says. “At the same time, if teams have to spend hours vetting and validating data, that makes it impossible for decision-makers to react at speed.” 

Despite the effort involved in getting predictive analytics up and running properly, the benefits are obvious to Petter. 

“I don’t think 2023 will be the year to leave any kind of chink in your corporate armour,” he says. “What predictive analytics enables is valuable to business leaders. With the insights it delivers, this technology has huge potential to turn data into gold.”