As companies are overwhelmed by torrents of data, they’re turning to automated tools that can dial down the noise and send clear signals to decision-makers.
In the midst of the Covid-19 crisis, even the world’s biggest businesses had to be able to pivot fast. For some, it was solutions powered by artificial intelligence (AI) that made swift decision-making possible, helping them be fast and nimble rather than slow and cumbersome.
When airlines began to close or limit routes in the frantic early days of the crisis, for example, the air freight market became chaotic. “The schedules and rates were changing every day,” says Alan Holland, CEO of Keelvar, which produces AI-powered strategic sourcing software. “It was a classic example of how AI could help when there was a deluge of data and urgent decision-making needed.”
Some of Keelvar’s pharmaceutical customers, who would usually take six months to a year to roll out any new software implementation, came looking for air freight sourcing bots that could be deployed within days. “These bots were able to get medical supplies into US hospitals on time,” says Holland.
Not only could AI-powered bots find the most efficient route through the supply chains disrupted by coronavirus, but they could also spot bad actors and over-inflated prices, and respond by rerouting goods through other airports.
The new normal is here to stay
For many businesses, there’s no going back, says Holland. Many of his customers that saw success with an initial handful of AI sourcing bots are now seeking to implement full automation programs across 30 or 40 categories.
This move to automation is coming across business processes and across industries, he adds. “Wherever people like spreadsheets, wherever you can be quantitative and measure things, automation is coming down the track.”
If you can’t measure it, you can’t manage it
Businesses worldwide are measuring more than they ever have, but they’re also generating more data than ever and often becoming overwhelmed. They seek to use data to better manage their business but then find they can’t manage the data.
“The key is not to gather more data,” says Holland. “It’s to manage and make sense of it. You need to decipher it, find patterns and problems, do diagnosis, find optimal outcomes, and understand trade-offs. That’s all manageable at a small scale, but quickly becomes impossible faced with torrents of data.”
In procurement, for example, companies may pay a data vendor for information on supply chain disruptions and another vendor for benchmark data to understand what competitors are paying in the market. Both would generate a wall of data, says Holland, likely overwhelming the procurement team.
“Companies are seeing an exponential increase in the amount of data but have the same number of people to deal with it,” he adds. “Those people are holding up a white flag and saying ‘Stop!’”
Time to call in digital reinforcements
The answer? Digital workers to help human workers, he says. Bots can do the heavy lifting of sifting through hundreds of megabytes of data, cleansing it, evaluating options, applying constraints as requested by the business, extracting concise information, and offering some options on the next best step for the business. The human dealing with it can then make a qualitative decision without having had to wade through raw data.
Data analysis that could take a person weeks to complete can be done in minutes by AI. “This saves 90% or 95% of people’s time for managing supplier relationships and strategic decision-making rather than tactical activities.”
Some industries are racing ahead when it comes to automated procurement, according to Holland. These include manufacturing, retail, medical devices, and industrial goods. What unites them? The fact that a higher proportion of their revenues goes back to suppliers for parts, ingredients, transport, or other services.
“Any industry like that has got to be good at procurement,” he says, “because it’s actually an existential threat to their business if they’re not. If their competitors are better at procurement, they could die.”
Automation will shine in lean times
As a global recession looms, it’s worth noting that industries often turn to AI-based sourcing tools when margins become tight. Holland has seen this in pharma, for example, as the rise of generics squeezed margins. “It’s easier for companies to increase profits by becoming better at procurement,” he points out.
The winds of AI will continue to bring seachange right across supply chain and procurement management over the next five to 10 years, he adds.
He expects AI to take hold of all quantitative aspects of procurement, not only measuring demand and establishing what needs to be bought, but also consolidating softer scores like performance levels, speed of suppliers and satisfaction levels.
The bots are coming, but they’re not taking over
Humans won’t be made redundant, however. “There will still be plenty of roles within procurement, but they will be focused on relationship building and innovation discovery, which is researching new products suppliers are developing that might benefit your business.
“Businesses will get better at measuring the softer benefits that good suppliers bring and good suppliers should welcome new tools that can measure all the value they provide. It will really help build mutual trust.”