There’s a growing range of AI and automation tools available to food and drink wholesalers, and plenty of reasons to be interested in what they can do. From improving stock forecasts to making better use of warehouse labour, the opportunities extend across much of the business.

The challenge is to know where to start. Most wholesalers already have a long list of things they’d like to improve, and there’s only so much time and money available to invest in new technology. An AI application might offer impressive capabilities, but that doesn’t automatically make it the best investment. What’s more useful is understanding where the greatest opportunities lie and how much difference they would make to the business.

Key Points

  • The best starting point for AI and automation is a business problem where improvement would make a meaningful commercial difference.

  • Potential savings need to be weighed against the cost and effort involved in achieving them.

  • Testing an application against clear performance measures helps wholesalers decide whether it’s worth a wider investment.

Where will AI and automation make the biggest difference?

Every wholesale business has processes that take more time or resources than they should. Some are obvious, such as manually entering customer orders. Others are harder to quantify, particularly where experienced staff have developed ways of working around inefficiencies.

Warehouse labour planning is a good example. Managers deal with the challenge of balancing the work coming through the door against staff levels every day. They know when the pressure tends to build, but changes in order volumes and picking requirements make it difficult to anticipate exactly how much labour will be needed.

AI offers the opportunity to analyse patterns in operational data and support more informed planning. Before exploring that possibility, though, it helps to understand what the existing approach is costing the business. How frequently is overtime needed at short notice? How much time do managers spend reorganising work when demand changes?

Similar questions apply to stock management. AI-powered demand forecasting offers the potential to improve purchasing decisions, but its commercial value depends on the problems the business is experiencing. Frequent stock shortages have a direct effect on customer service, while excess inventory ties up working capital. Understanding the extent of these problems helps establish what an improvement would be worth.

When there are several opportunities to consider, the priority shouldn’t necessarily be the task that looks easiest to automate. A relatively small improvement in an expensive or business-critical process can be worth considerably more than eliminating hours of low-value administration.

It’s also worth looking at how widely a problem affects the business. Improving order accuracy, for example, reduces the work involved in correcting mistakes and helps avoid problems further along the fulfilment process. An improvement that benefits several departments may justify greater attention than one that saves time in isolation.

Does the benefit justify the investment?

Once a worthwhile opportunity has been identified, the next question is what the business stands to gain and how much it will cost to achieve it. This is where the distinction between AI and conventional automation becomes important. Many repetitive tasks can already be handled through established rules and connected systems. Moving orders between applications or generating routine documents doesn’t necessarily require AI. Introducing more sophisticated technology where simpler automation will do the job can add expense without delivering any additional benefit.

AI is more relevant where the work involves interpreting complex information or identifying patterns that are difficult to manage through fixed rules. Demand forecasting and warehouse resource planning are potential examples, although the value depends on the quality of the information available and how the results are used.

There’s also a difference between saving time and reducing costs. If an automated process removes an hour of administration each day, that doesn’t necessarily translate into a direct financial saving. The business may instead gain capacity to handle more orders or give employees more time to deal with customers.

Those benefits are worthwhile, but they need to be assessed realistically. A projected saving is less useful if the time released is spread across several people in small amounts, or employees have to spend much of it checking the automated output. The investment itself extends beyond the price of the software. Integration and data preparation can add considerably to the initial cost, while training and ongoing support require further resources. Two applications offering similar potential improvements can therefore have very different business cases.

For wholesalers with limited budgets and competing priorities, a relatively straightforward automation project that delivers a reliable improvement may be a more sensible starting point than an ambitious AI implementation with a longer and less certain return.

What will it take to make it work?

A promising application can become considerably less attractive once the practical requirements are understood. Consider a tool designed to automate the processing of orders received by email. It can extract product details and quantities, reducing the need for manual entry. But if product information is inconsistent or customer records need correcting, much of the work remains. The time saved at the beginning of the process can be offset by the effort needed to resolve problems later.

This is one reason to assess the complete workflow before committing to a new tool. In a wholesale operation, information often moves between ERP and warehouse management systems, alongside ecommerce and other applications. The cost of connecting those systems and preparing the data can materially affect the return on investment.

People are equally important to that assessment. The employees doing the work understand the practical details that aren’t always apparent when a process is viewed from outside. Involving them helps establish whether a proposed solution addresses the real problem or simply creates additional work elsewhere.

Where AI supports decisions, employees also need to understand what its recommendations mean and when they should be questioned. A labour forecast, for example, can help a warehouse manager plan their resources, but it doesn’t remove the need to respond to unexpected changes or draw on their own experience.

These requirements need to be part of the business case from the outset, rather than issues to resolve after the technology has been purchased.

Proving the value before scaling up

Forecasting the benefits of a technology investment is one thing. Establishing whether those benefits have been achieved is another.

For many applications, a limited trial offers a practical way to find out. Rather than introducing a new tool across the business, a wholesaler can begin with one area of the operation and agree what improvement would justify expanding its use.

The starting position needs to be understood first. If the objective is to reduce the time spent on processing orders, that means knowing how long the work currently takes, including the effort involved in correcting errors. Warehouse labour applications can be assessed through overtime requirements and productivity, while demand forecasting can be measured against stock availability.

The measures should reflect the reason for making the investment. A tool that processes orders more quickly isn’t necessarily delivering value if it also increases the number of errors requiring attention. Likewise, a more accurate forecast needs to translate into better purchasing decisions before its commercial benefit becomes clear.

It’s also important to allow enough time to see how a solution performs under different operating conditions. A warehouse application tested during a relatively quiet period may behave differently when order volumes increase.

A trial provides an opportunity to identify unexpected costs and understand how staff use the technology. It also gives the business evidence to decide whether the results justify further investment. Sometimes the sensible decision will be to improve the application before expanding it. In other cases, the results may suggest that resources would be better directed elsewhere.

Exploring AI in warehouse operations

At BCP, we’re exploring the practical potential of AI through our Knowledge Transfer Partnership with Manchester Metropolitan University.

The two-year project brings together BCP’s knowledge of food and drink wholesale operations with academic expertise in artificial intelligence and data science. Its initial focus is on researching how AI can support labour and warehouse optimisation within BCP’s Accord warehouse management system.

Warehouse operations present an interesting opportunity because workloads change throughout the day, making resource planning difficult. Understanding patterns in operational data can help identify where better information would support more effective decisions.

The research is at an exploratory stage, rather than representing an existing AI feature within Accord. Its purpose is to investigate what technology can achieve and how potential improvements can be assessed in practice.

For wholesalers considering their own AI investments, that combination of technical capability and operational understanding is important. An application needs to work within the realities of a busy wholesale business, with benefits that can be demonstrated rather than simply anticipated.

Making AI investment work for your business

As the range of AI and automation tools grows, wholesalers have more choices about where to direct their technology budgets. The difficulty isn’t necessarily finding something that can be automated. It’s deciding which opportunities deserve investment ahead of the many other demands on the business.

That decision becomes easier when potential benefits are weighed against the cost of implementation and the changes needed to make the technology work. A modest, dependable gain in a critical process can be more valuable than a sophisticated application whose benefits remain uncertain.

Wholesalers don’t need to pursue every new capability to make progress. They need to identify the opportunities that are commercially worthwhile and build on results they can demonstrate.

If you’re reviewing where AI and automation can improve your wholesale operation, BCP can help you consider the opportunities in the context of your existing systems and processes.