Tech Check: Does Automation Matter More Than Generative AI?

Meghan Hall
·6 min read

Generative AI has been all the rage for retailers and brands—but the industries behind the scenes lack mirrored enthusiasm for the technology.

Two new data sets—one from Keelvar and another from Here Technologies and AWS—show that leaders in sourcing, procurement and logistics see the benefits of automation more readily than they do the potential use cases for generative AI systems.

More from Sourcing Journal

Sourcing may reap unique benefits from automation

According to Keelvar’s data, 42 percent of sourcing and procurement respondents indicated they will prioritize automation to free up time to focus on strategic efforts.

Using automation could be a boon to future success as 30 percent of respondents to Keelvar’s survey reported struggling with increasing demands in the workplace, said Alan Holland, the CEO of the Irish sourcing solutions company.

“If you look at [tasks like] supplier relationship management, innovation management and researching the market that you need experienced procurement professionals, they’re not getting sufficient time in their working days, and they’re bogged down in a lot of tactical work that machines are better placed to do,” he said. “Unfortunately, strategic work gets deprioritized in favor of tactical work when something is on fire.”

While sourcing professionals reported an interest in automation, they don’t have the same fire for generative AI as other industries do. Keelvar’s data shows 18 percent of those surveyed have interest in implementing generative AI in procurement.

Though the general, cross-industry sentiment has been that those not using AI will regret their decision later, Holland said generative AI may not be the answer to sourcing’s biggest problems.

Instead, he said, other subfields of AI could address real-world issues more effectively than large language models (LLMs)—at least for sourcing’s use cases.

“Just because you’ve got a hammer, not everything’s a nail. You’ve got to bring the whole AI toolbox to play here and choose the right tool for each task. Sourcing is a problem comprising multiple separate tasks, so you need different tools for each of the tasks,” Holland told Sourcing Journal.

One such example, he said, comes into play with supplier recommendations. While generative AI can provide recommendations, Holland said, Keelvar has observed incongruities between the information LLMs provide and the actualities of supplier relationships and information. Instead, he said, sourcing professionals should consider using recommender systems, a different type of AI, to help them make choices around suppliers.

That’s not to say generative AI systems have no place in sourcing—Holland said they can be useful for data aggregation and summary, as well as mapping out sourcing workflows. While procurement professionals don’t have big dreams for their generative AI implementation in 2024, two in five respondents noted that they have considered using generative AI for data analysis and predictive insights; one in three respondents said they had thought about using it for better user experience or automated content generation.

Though some use cases could prove beneficial, Holland said, on the whole sourcing professionals could have a chance to “benefit from the expensive learnings of others” at the moment where generative AI is concerned.

“There may be a second-mover advantage here,” he explained. “I’m very bullish on AI; I just recognize that generative AI is suffering from excessive hype that I hope doesn’t impact people’s perceptions of the wider field of AI and its usefulness.”

Transportation and logistics’ struggle against automation

Sourcing isn’t alone in its lagging attitude toward the implementation of AI systems. New data from Here Technologies also shows adoption of AI, data analytics and automation has been slow going in the UK and the U.S. among logistics companies.

Just 19 percent of UK transportation and logistics companies have deployed AI for decision making, tracking and predictive maintenance. In the U.S., 34 percent of T&L respondents noted that their company uses AI capabilities in its supply chain operations.

Incorporating AI and automation can prove more difficult without proper data infrastructure in place. Here Technologies’ data shows that only about half of transportation and logistics companies already have systems for data analytics in place.

“The study reveals that only 50 percent of T&L businesses across Germany, the UK and the U.S. use basic data analytics. While the U.S. still has room to grow in analytics and AI implementation, the country fares better than its European counterparts with the U.S. having the highest utilization rate at 63 percent, compared to the other two countries respondents at half or less,” noted Nina Hallquist, senior manager, market intelligence at Here Technologies.

For those companies struggling to implement technology one in four respondents said cost continued to be their biggest barrier; 12 percent said potential disruption to existing services deterred them and 11 percent said lack of internal expertise on the technology kept them from adopting it.

Integrating automation and sustainability

While many companies haven’t taken the leap to automate their processes, Holland noted that doing so could help revolutionize their approaches to a major problem facing most industries right now: sustainability.

Per Keelvar’s data, 60 percent of sourcing and procurement companies will prioritize sustainability this year; the data does not reflect what percentage of those companies have begun working toward those goals. In transportation and logistics, the reality proves much more bleak. According to Here Technologies, 62 percent of those surveyed said their organizations had no goals or metrics around sustainability in place.

Hallquist said using automation and driver routing technologies could help companies looking to achieve sustainability-related goals.

“Automated fleet tour planning and driver routing leverage a vast amount of data, such as vehicle restrictions and real-time traffic data, to increase efficiency and reduce fuel consumption. Companies that already use these technologies can go one step further and aim at reducing truck idling and dwell time. Less idle time means less fuel waste and emissions, in addition to happier drivers and customers,” Hallquist said.

As 2030 creeps closer and companies consider commitments around emissions and environmental concerns, Holland agreed that automation could be the missing piece to help companies put their ideas in motion.

“Most [companies] haven’t made the connection between automation and sustainability. They feel that operationalizing sustainability is going to be time consuming, [and] they don’t have the personnel to do an additional job on top of what they’re doing,” Holland said. “Those that are more cutting edge recognize that automation is perfect for bringing in additional data points… so you can have a richer and richer decision-making framework.”