Need a Crystal Ball?

By Gian Schiava

May 2023

How to predict warehouse workforce demand

What can labour management software (LMS) do for warehouse operations? Where are its flaws? What is its future? And how can you keep your employees motivated when their tasks are increasingly being assigned by software? Gian Schiava investigates the challenges for LMS and the role of machine learning in their solution.

One fundamental issue for managers is that it’s not always clear what labour management software should enable you to do.

With the rise of e-commerce, the growing number of sales channels and the ever-increasing pressure from deadlines, supply chain professionals need their organisation to be more responsive than ever. Internet orders can be placed at any time and need to be fulfilled quickly, sometimes in just a few hours.

To organise activities efficiently, today’s supervisors rely on a full-scale WMS (warehouse management system) to define the priorities. A WMS, however, focuses primarily on your products and the available systems, and to a much lesser degree on your workforce. This is where labour management software (LMS) comes into play.

Not such a new tool really

When digging into this subject, we discover that the first labour management software solutions entered the market more than a decade ago. Many WMS suppliers have been including an LMS tool within their offering since then. Whilst managing workforces in production environments had been common, fine-tuning human resources to warehouse systems and order output appeared to be quite a different ball game.

In earlier times, logistics professionals reacted to labour needs on a daily basis. Armed with data from the ERP (enterprise resource planning) system (or the sales department) and an Excel sheet, they tried to anticipate demand as closely as possible. This involved converting ‘raw’ data into actual workload figures. In many cases, temporary workers would be hired, hoping the estimated workload proved to be right.


Too many workers in the warehouse or too few? Traditionally, it’s been very difficult to anticipate changing daily needs.

Warehouse challenges

One fundamental issue for managers is that it’s not always clear what labour management software should enable you to do. There are many fancy marketing descriptions, but Dutch-based IPL Consultants offers this clear overview: “Automated tools help warehouse managers to calculate how much manpower, with which specific skills, is needed to fulfil certain tasks. The challenge is that, contrary to production environments, order lead times in a warehouse environment are very short.”

This time restriction may well be a major reason why labour management tools still haven’t been adopted on a large scale.

On further investigation, we find several other reasons for this low adoption. Our source is Grégory Lecaignard, Software Product Manager at SAVOYE, a provider of intralogistics systems and software. This French specialist’s extensive research has revealed five fundamental objections to LMS adoption. Many logistics professionals considered today’s solutions to be:

  • Short-sighted. The systems’ provision of workload visibility often relates only to today. There is no long-term overview.
  • Continuously hungry for data. Often you need to input the complete schedule for all operators, along with their names. And tomorrow, the same again. This requires too much time.
  • Self-centred. They cannot analyse the performance of any processes that are not controlled by the WMS. However, on a daily basis, there are many important tasks to be carried out which are not tracked by that information system, like pallet inspection or value-added services.
  • Imprecise. This is often caused by differences in vision between the supplier of the solution and the actual operator.
  • Unclear on ROI. Given the four flaws above, it’s impossible to calculate a return on investment.
Gregory Lecaignard_1

Grégory Lecaignard, Software Product Manager, SAVOYE

Machine learning is the key

According to Grégory Lecaignard, automated tools for labour management will only really start to take off with the insertion of machine learning algorithms. Only then will the LMS be able to make proper forecasts on which the manager can base timely anticipation of the human resources needed. Without that information, you will always have to cope with either the expense of too many people in the warehouse or, even worse, the delivery delays resulting from having too few.

Additionally, a system capable of learning not only does the planning but manages the assignment of all the tasks to the right people, taking into account each worker’s skills, experience and other factors.

Grégory concludes: “Labour management should be a key function of a distribution centre. But without forecasts, such a system is completely blind. Forecasts sometimes exist in another system, but data are expressed as sales units and never in terms of expected workload. Only a machine-learning-based forecasting system, as part of the workforce management system itself, will produce reliable and detailed results that the manager can use to determine how many workers he needs to hire for the next cycle. A good machine learning system with enough historical data (more than 15 months) can look forward up to 30 days and can produce a result with only 5-10% error margin.”


LMS combined with machine learning will enable accurate forecasting of warehouse labour needs.

Then add motivation

For the manager there are additional advantages, aside from the perfect coordination of people and resources. He or she will gain insights into output, per channel, per activity and even on an individual basis. The system could therefore also be used to introduce output-based payments – but given the scarcity of labour these days, that could well be the wrong way forward.

Instead of using the new insights from LMS as a stick, it’s better to use them in a form of encouragement called gamification. The idea of this trend is to increase employees’ productivity by motivating them and making their tasks more playful. For example, game elements can be introduced to promote specific objectives, and can be applied both to individuals and to teams.

The combination of gaming and labour management can have multiple benefits – and not just in terms of much better resource planning and cost savings. It can lead to increased involvement, transparency and, perhaps best of all, keeping and developing a loyal and skilful workforce.