15 Sep Warehouse KPIs: what works and what does not

What research says about performance indicators on the shopfloor

Almost every warehouse works with KPIs. Productivity per hour, picking errors, order lead time. But measuring performance indicators changes nothing in itself. What matters is what management and operators do with the figures.

With Sander de Leeuw, professor at Wageningen University, I investigated which ways of working with performance indicators actually deliver results. More than a hundred companies completed the survey. Of the twenty practices we found in the literature, seventeen turned out to be statistically related to behaviour that leads to better performance. The study was published in 2011 in the scientific journal “Journal of Operations Management”.

What KPIs deliver

Of the managers surveyed, 71% saw higher productivity, 73% improved quality and 37% faster inventory turnover as a direct result of performance indicators. In a substantial share of cases the improvement exceeded 5%.

The interesting part is the combination. In practice you usually see a trade-off: higher quality comes at the expense of productivity, or the other way round. Companies with well-functioning performance indicators achieved improvements on several objectives at once. They managed to break that trade-off.

Three effects that make the difference

The improvement does not come from measuring, but from the behaviour that measuring triggers. We found three clusters of behaviour that occur largely independently of each other and all three relate to better performance.

Understanding. Management and the shopfloor have a representative and up-to-date picture of the performance delivered. They know the company objectives and understand how their work contributes to them. Bottlenecks and their underlying causes are quickly visible.

Motivation. Operators see the indicators as a means to improve. They are motivated to score well, and experience that good performance genuinely contributes to the business result. That last point requires the objectives of different departments to be aligned.

Focus on improvement. Operators understand the relationships between indicators and business results, and can make independent decisions that benefit the whole. They are trained in interpreting management information and analysing causes. The indicators are embedded in an improvement cycle.

Defining KPIs

Use an existing model. Companies working with the Balanced Scorecard, EFQM, SCOR or a corporate standard score better than companies that design their own measurement model. Existing models have a clear structure and provide a complete and balanced picture.

Derive them from your objectives. Operational indicators that follow from strategic and tactical goals relate to all three behavioural clusters. This is one of the strongest relationships in the study.

Cover efficiency, effectiveness and flexibility. Only 15% of companies do this, while it clearly relates to better understanding on the shopfloor.

Define them jointly with all departments involved. Not each department separately. Aligned indicators prevent departments from working against each other.

Use objective standards. Standards based on customer requirements, benchmarks, time studies or historical data work better than estimates by management or operators. With an objective standard, the indicator genuinely tells you whether performance is good.

Implementing KPIs

Do it to improve performance. Cutting costs or raising service levels, not to comply with legislation or to assess people. That sounds obvious, but the motive carries through into how people experience the system.

Involve team leaders and operators. Making them part of the implementation team affects all three behavioural clusters. Only 37% do this.

Pay attention to cultural change and training. Performance gets assessed concretely, operators have to learn to see connections and think along about improvements. Companies that actively invest in this are more successful.

Reporting KPIs

Calculate automatically. A predefined computational model in Excel, Access or a BI system works better than manual calculation or the standard reports from ERP, TMS or WMS.

Make the underlying details electronically available. Besides the indicators themselves you need the details to trace causes. Dashboards with direct drill-down work best.

Report to the shopfloor daily. This relates to all three behavioural clusters, yet only 23% of companies do it. Current figures speak to the imagination more, because people can still remember the causes.

Using KPIs

Set goals at individual or team level. Not at department or company level. This is one of the strongest relationships in the study: people steer on what they can influence. Goals at department or company level remain too remote to act on.

Use a fixed improvement methodology. An approach that ties actions to measurements, such as the Deming cycle, relates strongly to focused improvement. Only 28% apply this.

Discuss the results between departments. Structurally, not incidentally. This affects all three behavioural clusters.

Lead on task and relationship equally. A leadership style that gives both equal attention works better than a one-sided one.

The controversy: measuring individuals

Reporting individual performance meets resistance. It is often associated with a system that monitors people constantly.

Yet reporting individual performance — publicly or confidentially — relates to more improvement-focused behaviour. Making it public, for instance with a list on the wall, produces no demonstrable increase in social pressure. It does give operators a better overview of total performance, so they think along more actively.

Measuring is something different from assessing, though. Operators usually work in teams and depend on each other. Goals at team or individual level work; goals at department or company level do not.

What makes no difference

A third of companies link performance to rewards or salary increases. That practice shows no significant relationship with any of the three behavioural clusters, and there is no difference in improvement between companies with and without performance-related pay.

The same holds for a second practice: a separate staff function that gathers the data and computes the indicators. Of the nineteen practices tested, these were the only two that dropped out.

Useful to know for anyone considering a bonus scheme to get performance indicators off the ground. The energy sits elsewhere.

How many practices should you apply?

One of the most striking findings is not about which practices you choose, but how many. The more you apply, the higher the improvement achieved — regardless of the combination.

That makes the advice simpler than it seems. You do not have to work out first which approach suits you best. Start with a few, and expand.

On average a company applied eight, with a spread from zero to fourteen. So there is room almost everywhere.

What it comes down to

KPIs improve shopfloor performance considerably. That improvement comes from what people do with the figures: understanding where they stand, becoming motivated to score better, and learning to see the connections between their work and the result.

That calls for a well-considered model, objective standards, fast and detailed feedback, and goals at the level where people actually have influence.

De Leeuw, S. and J.P. van den Berg (2011). Improving operational performance by influencing shopfloor behavior via performance management practices. Journal of Operations Management 29(3), 224-235. doi.org/10.1016/j.jom.2010.12.009 — I will send you the full article on request.

About the author
Jeroen van den Berg is the author of Highly Competitive Warehouse Management and holds a PhD in warehouse algorithms from the University of Twente. He has been advising companies on warehouse optimization and WMS since 1997, running his own consultancy since 2001. He develops Metrica, a Warehouse Optimization System.