The softer conditions of the recent past have not stopped investments in technology for companies in the metals space, even if one major pullback is taking place. That was one of the conclusions drawn from the most recent Technology in Metals survey conducted by Crowe.
Completed each year for the past 13, the survey asks metals participants throughout the supply chain, here and abroad, about their technology investment plans. Though Crowe customers are included among the respondents, the vast majority are not currently using Crowe’s software.
In the most recent survey, 58 percent of respondents said they intend to increase investments, up from 51 percent in the previous survey.
But that doesn’t tell the entire story. For the first time in many years, there was a drop in the number of service companies implementing new ERP systems.
This shouldn’t have been terribly surprising, says Tony Barnes, principal and Microsoft Cloud Solutions leader for Crowe. With companies investing so consistently for a decade, there was bound to be a period of pullback.
“While companies are continuing to invest, there’s a little bit of digital transformation fatigue or technology investment fatigue. Some of it from a cost perspective. Some of it from a change perspective. IT teams, e-business users, have had to manage so much change. They need to pause, step back and catch their breath.”
One other noteworthy development, Barnes says, was around the rate of adoption of AI. The survey approached the subject in several different ways, asking companies whether they’re using it, if they plan to and, if not, why?
Through the first quarter of 2024, the answer to whether they were using it was typically, “No.” More than half of the companies surveyed earlier in 2024 said they were not using it and had no immediate plans to. Reasons cited for their unwillingness to adopt AI was not knowing enough about it or not having the resources to support it.
“Some of it is generational, some of it is unwillingness to try new technology. How do we get everybody comfortable and establish expectations of it?” he says of the reluctance. However, he also predicts, “I am confident in saying that number will drop (of people not planning to use it) in a major way in 2025.”
He says his company is already talking to its customers about how to tackle AI in a manageable and efficient way.
First, he says, it should and will start small. Rather than implementing some major program, companies should begin by taking advantage of the AI capabilities within existing software programs used every day.
“If I use Outlook, if I use Teams or some other chat service, those major services are all embedding AI in their applications today. We tell companies before they go out and get some third party AI solution, look at their email inbox today. Look at the tools you use for meetings and see what AI solutions they offer for transcripts, simple things you can do to make your meetings more efficient.
“It’s not just two people in a meeting that get that information. The opportunity to share that information and everybody benefiting will transform certain functions within this industry,” he adds.
Of course, adoption is just the first part. Like any technology, it’s only useful if the information it generates produces real results. “You have to turn it on, but you have to educate folks on right-use cases and wrong-use cases. How does it benefit us?” Barnes says. “Turning it on is the first step and continuous education and monitoring of that is the second step.”
Monitoring is monumentally important. You have to have the right controls around the data being used, he warns. If you take proprietary data, whether around pricing or products, into a publicly available AI model, that tool now has access to that data to be used to retrain for someone else.
“If you get an RFP from a customer that’s confidential and your sales team plugs that into ChatGPT and their confidential information is in there, you’ve opened yourself up to some risk.”