Argon & Co explains why logistics businesses need to start with value, operating models and data foundations before investing in AI and automation.
As artificial intelligence and automation become more common across logistics operations, the challenge for businesses is no longer whether to invest in new technology. It is how to make those investments deliver measurable value.
For Argon & Co, the starting point is not the technology itself, but the business problem it is intended to solve. Logistics leaders need to understand where value is being created, where it is being lost, and whether automation or AI is the right mechanism to address the issue.
Cameron Austin, Managing Principal at Argon & Co, says the strongest performers are those taking a pragmatic approach to technology adoption.
“I think where industry leaders are getting it right is that they have a balanced and pragmatic view,” says Cameron. “You must be pro-technology but not get caught up in the hype.”
That distinction is important. Automation can accelerate a process, but it does not necessarily improve it. If the underlying workflow is poorly designed, automation may simply increase the speed at which inefficiency moves through the business.
For logistics operations, this means investment decisions need to be tied to clearly defined outcomes. These may include reducing cost to serve, improving labour efficiency, lowering cost per unit, optimising transport spend, improving margins, increasing on-time in-full performance, or improving order accuracy.
Bart Gill, Associate Partner at Argon & Co, says logistics value needs to be measured through business outcomes, rather than activity metrics.
“Real logistics value is about measurable business outcomes, not just activity metrics,” says Bart. “It’s about improving your cost, your service and your resilience, and doing that simultaneously.”
Once those outcomes are defined, businesses can begin assessing demand patterns, order profiles, market volatility, customer expectations and the operating model needed to support the desired result. This allows leaders to understand where automation can remove constraints, improve decision-making or strengthen operational performance.
Bart says this groundwork is essential before moving into process design or technology selection.
“Start with what the business outcomes are and make sure they are clearly defined,” says Bart. “Examples would be service levels, cost reduction and scalability, and then translating those into measurable targets.”
AI is also creating new opportunities in parts of logistics that have traditionally been difficult to automate. While warehouse management systems, transport management systems and other platforms already support structured workflows, many operational decisions still rely on unstructured information, including emails, status updates, exception reports and fragmented data sources.
Marjan Torshizi, Managing Principal at Argon & Co, says this is where AI can help connect information and support more cognitive forms of automation.
“If we look at WMS, TMS and other systems, AI is bridging the gap that exists in an unstructured world,” says Marjan. “It can elevate automation to more of a cognitive workflow automation.”
Rather than simply executing predefined tasks, AI can help gather information, interpret context and determine what needs to happen next. In logistics environments, this can support better exception management, faster decision-making and more integrated workflows between systems.
Cameron says logistics businesses should view automation across three connected categories: physical automation, process automation and decision automation. Physical automation may include robotics, conveyors, sortation systems or automated materials handling equipment. Process and decision automation, meanwhile, may be enabled through AI, data platforms and system integration.
The greatest value often comes when these forms of automation are considered together. When deployed in isolation, they can create new bottlenecks or disconnected pockets of capability.
“Businesses should think about the different forms of automation holistically,” says Cameron. “Each of them has different benefits, but the greatest impact comes when they are integrated.”
A strong business case also depends on accurate baselines. Businesses need to understand their current performance in detail, then compare it with a realistic future-state model. That includes a full view of costs, benefits, implementation requirements, change management effort, adoption assumptions, risk-adjusted savings and accountability across the organisation.
Cameron says this helps businesses assess whether a solution solves the right problem, fits the operation and delivers measurable value.
“The most important thing for a successful business case is accurate baselines,” says Cameron. “That means comparing the current situation and the updated automation baseline on a like-for-like basis.”
However, even a strong technology business case will fall short if the operating model is not redesigned around it. For Argon & Co, the operating model is the link between deploying new capability and realising value from it.
Automation changes how decisions are made, how work is allocated and how people interact with systems. It can also change the relationship between functions that may have previously operated in silos.
Bart says this is why businesses need to treat automation as an operational transformation, rather than a technology project.
“Technology alone creates capability, but the operating model determines how that capability will be embedded into decisions, processes and behaviours,” says Bart.
This also changes the role of the workforce. Automation does not remove the need for people in logistics operations, but it shifts where their expertise is applied. Employees may spend less time on manual execution and more time managing exceptions, monitoring performance, interpreting data and orchestrating flows across automated systems.
That shift increases the need for hybrid capabilities, combining operational understanding with digital literacy, problem-solving and data interpretation. Maintenance, technical support and reliability roles also become more important, particularly in automated environments where uptime is central to performance.
Data quality is another critical foundation. In manual operations, people often compensate for incomplete, inconsistent or poorly structured data. In automated environments, those issues become harder to hide and more expensive to correct.
Marjan says businesses need to understand how their data is defined, tagged, classified and governed before expecting AI to deliver reliable results. This includes understanding what information is confidential, sensitive or usable across the organisation, as well as how data moves between systems.
“Traditionally, in manual operations, people can often cover up the data issues and plug the gaps,” says Marjan. “But as soon as you move the process into an automated flow, data issues can make AI implementation very expensive.”
Without proper integration into the existing system landscape, AI solutions can also struggle to deliver the efficiencies expected. A standalone tool may provide localised benefit, but it will not necessarily improve the end-to-end flow of work.
For logistics leaders considering AI or automation investment over the next 12 to 24 months, Argon & Co’s advice is to start with the problem, quantify the value, assess data and organisational readiness, and redesign the operating model where needed before selecting technology.
“Clearly define your business issue, quantify its value, assess your data and organisational readiness, redesign the operating model where necessary, and only then select the right technology,” he says. “Let strategy drive technology, not the other way around.”
Marjan says this also requires a more coordinated approach to AI strategy.
“It cannot be a technology-led initiative,” she says. “It has to be a top-down mandate at the whole enterprise level.”
Ultimately, the organisations that succeed with AI and automation will not necessarily be those that deploy the most systems. They will be the ones that understand where decisions are made, how work gets done and how technology can support a better operating model.
As Bart says, “It’s not the businesses that deploy the most technology or the most automation. It’s the ones that redesign based on where decisions are made and how work gets done.”
.png){kind=small}
