Fernando Alarcón, who works in procurement at Hapag-Lloyd, told delegates at TOC Americas in Cartagena that he would spend a hypothetical unlimited crisis-proofing budget on ports, warehouses and other physical infrastructure rather than on software or artificial intelligence, arguing that technology is turning into a commodity almost as fast as companies can buy it.
Alarcón was one of six supply chain and retail executives put through the exercise by moderator Miguel Garín, of Fundación ValenciaPort. Each was handed an imaginary unlimited budget to protect their company against future uncertainty, on condition that they spend it entirely on one of two things: information systems, algorithms and artificial intelligence, or physical assets such as vessel space, warehouses and cranes.
The panel had already reached something close to a consensus on a related question. Asked earlier whether technology or people should lead in a crisis, Elsa Muñoz, of AliExpress, the e-commerce arm of Alibaba Group, said artificial intelligence could not yet replace a manager's judgement. “Decision-making is something artificial intelligence still cannot do … it is an extraordinary tool,” she said. Alarcón agreed. “The answer is people, the people who are in charge of operations,” he told the panel, saying Hapag-Lloyd used artificial intelligence to model vessel scheduling around canal restrictions caused by drought but left the final decision to staff.
The volatility behind that caution was not abstract. Carlos Palacios, of GEODIS Colombia, said rates on one lane out of the Far East climbed from $2,000 to $4,000, then $5,000, before peaking at $9,000 and settling around $6,500, as clients delayed booking space in the hope that prices would fall. Milena Jiménez, of Grupo CargOK, described a 40ft container split into two 20ft boxes after a shipping line ran short of space and equipment: one half reached Colombia on 10 August, the day of an earthquake, and Buenaventura could not handle it for the next 30 days, she said. “The container, of course, could not be unloaded,” she said.
Hard assets over software
That consensus on people did not survive the unlimited-budget question intact. When Garín put the choice to the panel, Alarcón sided with infrastructure, arguing that technology's pace of change makes it a poor long-term bet. “Technology is advancing so fast … it is ending up becoming a commodity … I would invest in energy, in infrastructure,” he said.
Jiménez agreed, though her target was narrower: the containers themselves. “I would invest first in replacing these metal boxes we call containers, because we owe it to ourselves to renew that fleet,” she said, adding that carriers do not even own most of the boxes they move. Palacios also chose infrastructure, pointing to Colombia's underused access to Venezuela and the lack of international logistics distribution centres at its ports. “I think the first thing … would be to invest in infrastructure. Technology is fundamental once you have great infrastructure, but you can only invest in one of the two,” he said. Xavier Montes, of discount retailer Dollar City, which operates across Colombia, Central America, Peru and now Mexico, made the same choice, pointing to the new Chancay port in Peru and recent port investment in El Salvador as evidence of a wider regional infrastructure gap.
The case for algorithms
Not everyone agreed. Muñoz chose software, arguing that an e-commerce platform lives or dies on the algorithms managing customs data and delivery promises rather than on the ports themselves. Gonzalo Córdova, of Falabella Group, made the clearest case for the other side. “They say information is power … if I have information, along with the best algorithms, I think you can make better decisions and really be shielded from a crisis,” he said, describing plans to put GPS trackers on packages across the group.
Predict or react
Garín then put the panel through a second version of the dilemma: as newly appointed innovation directors, should they spend on anticipating the next crisis or on reacting to it faster? Here Alarcón and Córdova, on opposite sides of the technology question, found themselves agreeing. “We are never going to be able to anticipate everything … there is always a part you have to leave for the reaction,” Alarcón said. Córdova went further, casting doubt on whether prediction is realistic for the region yet. “We are not prepared, at least in my part of the business, to operate using predictive models. We are a long way from being able to operate on predictions,” he said.
Palacios took the opposite view. “The answer is to anticipate,” he said, arguing that Colombian companies rely too heavily on historical demand models and need predictive planning instead.
Muñoz closed the panel by siding with the reactive camp, against her own earlier vote for software. “I focus on the here and now, on what I have, and on the reactive,” she said.