Rethinking AI as a Solution to Everything
No matter where you look, headlines announce that the use of AI is shrinking the workforce. But implementing AI is not a workforce reduction strategy, says Rachelle Butler, director of Redwood Advisory at Chicago-based Redwood Logistics.
“I would never recommend to a shipper client that they use head-count reduction as anything more than a soft savings because you can't necessarily quantify head-count reduction, especially in relation to AI.” She acknowledges there are exceptions to every rule.
In general, Butler says, her clients want to know about opportunities when implementing AI and technology. Overemphasizing labor savings can distract from such benefits as improving transportation costs, service levels and operational decision-making.
AI does offer opportunities to shippers, logistics companies and supply chain organizations. It can be used to strategically solve business challenges. But is it the best tool for the job?
What Is Your Business Challenge?
When you were a child, your parents might have told you, “Stop jumping to conclusions.” In business, that translates to: “Don’t immediately jump to a solution.” There has to be something to solve before a solution can be developed.
Many organizations implement AI without having a business challenge to solve, Butler says. Or the problem is such a small segment of their business that it doesn’t warrant the investment, at least yet, she says.
“It goes back to that discipline of: Always identify and understand the business problem as well as the upstream and downstream impact,” she says. “What are you experiencing? And if you were to prioritize investing in solving that business problem, what is the metric you’re using? What is the ROI? You have to have some sort of baseline.
She continues: “We hear ‘AI’ all the time, so everybody’s just inserting AI” as the solution to all issues.
Butler recommends taking AI completely out of the equation when evaluating whether you need technology as a solution. “The whole concept of how you approach investment should not be thrown out the window just because you consider AI to be a cool piece of technology,” she says.
Establishing a Baseline
Before investing in technology, companies should first identify a specific business problem. Most likely, that problem will pertain to one of two things: cost control and service improvement, Butler says.
Industries may have different priorities when it comes to cost control. “But improving service across the board, industry agnostic, seems to be the No. 1 priority — and it should be,” she says. “How do you retain and increase revenue? You have a great service offering and a reliable service offering.”
To determine the problem to improve, Butler recommends creating a concept board that lists all the challenges the company is facing. “Consider the business and organization strategy. What is your own department’s strategy? How are your challenges correlated to that?” she says.
Next, create a heat map. Determine what could be quantified and the metrics that can be used. Establish a baseline.
“This usually helps you prioritize what you can realistically execute on, what can be measured and what improvement would look like,” Butler says. In logistics operations, for example, better and more strategic management of market conditions and disruption can be valuable to address.
Once these factors are identified, solutioning can start.
Solve for the Business Problem
Solutioning doesn’t mean automatically assuming AI is the right technology or strategy to address the issue.
But the decision can involve determining whether AI offers an advantage over existing technology, optimization or traditional automation. It means ensuring outcomes can be measured or quantified.
“It all comes down to if it actually addressing the problem, and if the implementation and the long-term maintenance, including governance, outweigh the cost of any alternative model,” Butler says. “I say this because I’ve seen examples where companies think they want to solve an issue with AI, when in reality, there are existing technologies that are far less complex to maintain and govern long term that solve for the business problem.”
When AI is chosen as the solution, user adoption is key. Even solutions with quantifiable benefits won’t deliver value when employees cannot or will not use them, Butler says. Sometimes there can be employee pushback. Some personnel may be reluctant to try new technologies or want to keep doing what they’ve always done.
“User adoption is one of the biggest challenges that clients are faced with post-implementation,” Butler says. Employee training, governance standards and structure on how to use AI on the job are paramount.
As with older technologies, AI will likely eventually become the norm. “But how it’s going to be used as an enterprise scale,” Butler says, “I think is in its infancy.”