Proximo Spirits Expanding Use of Decision Intelligence Across the Globe
Proximo Spirits is extending its use of decision intelligence within demand planning across 85 countries after successful pilots in the United States.
The company, which is a subsidiary of Becle and counts brands like Jose Cuervo, Three Olives, and 1800 in its portfolio, began leveraging the technology less than a year ago in the United States, its largest tequila market. The proof of concept started bearing fruit by learning demand patterns at the state level after six months, and the company has reduced forecast error down to 25%, according to Luis Gonzalez, Becle global supply chain director.
Beginning next month, Proximo will extend the technology to 85 counties across the EMEA, APAC and Latin regions. They also intend to expand the use of decision intelligence across their supply planning capabilities for integrated business planning as part of the second phase.
Change Management and Trust in AI
The spirits business is both seasonal and volatile, Gonzalez shared during a virtual event with Aera Technology, the decision intelligence solution provider it’s partnered with. Combined with noise caused by pandemic in the last 2-3 years, the market faces a combination of complexity, particularly in the United States.
Proximo has a lean team of just four people dedicated to producing forecasts and creating consensus demand plans. As a result of the decision intelligence platform, they can now only focus on exceptions and make decisions that the tool is unable to, said Gonzalez.
While much of the change management has been relatively easy because employees now have more time to make decisions in other areas of the business, learning to trust that the AI-infused platform is making the right decisions has a steeper learning curve, he said.
“In some cases you don’t understand what’s behind the decision, but when you see why the tool suggested some numbers, you realize it was correct,” said Gonzalez. “Sometimes it’s difficult for us as humans to understand the logic of what’s behind the process. However, if you trust the logic and the data, you can be sure the decision will be the right one.”
As a result, having the right quality of data to start with is absolutely critical, he noted, no matter the platform you use.
“Never underestimate the dramatic importance the master data has. … If you don’t have the right quality of your data, or the data is not the right structure, you will have to work it out in order for the tool to provide the right outputs,” he stressed.
“You have to have the right set of data so the tool will provide you the right value. Otherwise you will keep investing lots and lots of time trying to understand while the tool is providing the data, from decisions in some cases, it will not provide the right information.”
