How We Improved Warehouse Picking Accuracy and Reduced Picking Errors in 12 Months

Most articles about reducing warehouse picking errors and improving picking accuracy talk in abstractions. Optimise your slotting. Improve your pick paths. Invest in the right technology.

What they rarely show is the actual before and after. The specific numbers. The decisions that worked and the ones that didn't.

This is that article.

Over a 12-month period, a UK pharmaceutical logistics operation managing over 300,000 units dispatched daily to hundreds of pharmacy locations achieved a 21% year-on-year improvement in lines picked per hour (LPH) — the core warehouse KPI for any pick operation — driven by the same process changes that also reduced warehouse picking errors. No new equipment. No capital spend. Just better use of existing warehouse reporting tools, more deliberate operator coaching, and a clearer focus on warehouse picking accuracy.

This is how we did it.

LPH and Picking Accuracy: The Two Warehouse KPIs That Matter Most

Before covering what changed, it's worth explaining why lines picked per hour matters more than most warehouse KPIs that operations managers track — and how it relates to picking accuracy.

LPH is honest. It strips out variables like order complexity, product mix and staffing levels and gives you a single number that tells you how productive your warehouse picking operation actually is on any given day.

Total units picked is misleading — a day with lots of single-line orders looks very different to a day heavy with multi-line picks. LPH normalises for that. It tells you how efficiently your team is converting active time into output, and it's the clearest measure of warehouse picking efficiency available to most operations managers without specialist analytics tools.

It is also the metric your team can directly influence. Unlike fill rates or despatch accuracy — which depend on upstream factors — LPH is a direct measure of operator effort, process design and working environment. If you want to improve warehouse performance, this is where to start.

Importantly, LPH and warehouse picking accuracy are not in opposition. In this operation, the same operator behaviours that drove speed improvements — deliberate pick path discipline, controlled scanner technique, fewer interruptions — were the same behaviours that reduced picking errors. An operation that picks fast but picks wrong gains nothing on order accuracy. The goal is always both.

If LPH is going up, something is working. If it is going down, something needs attention. The same applies to picking accuracy.

Where We Started: Diagnosing the Causes of Picking Errors and Productivity Loss

At the start of the measurement period, the warehouse picking operation was running at a team average LPH that was functional but not exceptional. The distribution was wide — meaning there was a significant gap between top performers and the rest of the team.

That gap is actually a useful diagnostic. A wide LPH distribution tells you that the process is not the constraint. If process were the bottleneck, everyone would cluster around the same rate. A wide spread means individual behaviour, approach and habit are the variables — which is exactly the kind of problem you can solve without capital investment. It also tells you that warehouse picking errors and wasted motion are concentrated in specific operators and shifts, not systemic — which changes how you approach the fix.

In most pick operations, warehouse picking errors cluster around the same root causes: operators rushing through scanner confirmations to maintain pace, inconsistent pick paths that create location confusion, interruptions during peak pick windows, and a lack of real-time feedback on individual accuracy. The data in this operation pointed to the first three as primary contributors — all of them addressable through process change and warehouse coaching rather than technology.

Three things stood out from the initial data analysis:

• Top performers were not being studied or replicated
• Shift patterns were not being tracked against output
• Station-level data was available but not being acted on

These three gaps became the focus of the warehouse process improvement work.

What We Changed

1. We Started Measuring What We Were Ignoring

The WMS was generating daily pick reports but nobody was aggregating them systematically. Warehouse reporting tools were in place. Insight was not.

The first step was building a consistent measurement framework — pulling daily LPH data per operator, per shift and per station, and tracking it over time rather than looking at single-day snapshots. If your WMS also records pick errors or mispick events by operator, pull that data alongside LPH. The two together tell a far more complete story than either metric alone. This is the foundation of any serious warehouse continuous improvement effort: you need a baseline before you can measure change.

This alone changed the conversation. When team leaders could see weekly and monthly trends rather than yesterday's numbers, the discussions became more strategic and less reactive. Problems that had been invisible in single-day snapshots — like one shift consistently generating more warehouse picking errors than another — became obvious in the trend data.

Tools like Smartsheet can automate this kind of warehouse analytics and data aggregation if your WMS exports to CSV or Excel, saving significant manual processing time each week and giving you live warehouse reporting without additional infrastructure.

2. We Analysed Shift Performance Separately

One of the most actionable findings came from splitting LPH data by shift rather than looking at the team as a whole.

The late shift was consistently outperforming early shift on LPH. Not marginally — consistently and significantly.

This raised an immediate question: why?

The answer, after investigating, came down to a combination of factors. Late shift had developed stronger peer accountability norms. Pick station allocation was more consistent on late shift. And crucially, late shift had less management intervention during peak pick periods — operators were less interrupted.

This finding directly informed how we structured early shift operations. Reducing unnecessary interruptions during peak pick windows and tightening station allocation consistency drove measurable improvement within weeks.

3. We Identified and Replicated Top Performer Behaviours

The top performers — consistently delivering 40% or more above team average — were not working harder. They were working differently. And critically, their warehouse picking accuracy was better too.

Observation time with top performers revealed consistent patterns that were not being taught or replicated across the team:

• More deliberate pick path discipline within their zone — fewer wasted steps, fewer missed locations, and fewer pick errors caused by location confusion
• Controlled scanner handling and confirmation rhythm — they confirmed picks deliberately rather than rushing, which directly reduced picking errors caused by mis-scans
• Better management of their own replenishment timing to avoid dead time waiting for stock

None of this required new equipment or systems. It required observation, documentation and deliberate warehouse coaching.

We used a structured buddy system pairing mid-performers with top performers for focused observation shifts. The LPH uplift in the mid-performer group was measurable within the first month — and their operator accuracy improved alongside their speed, which is the pattern you expect when picking errors are driven by technique rather than pressure.

SafetyCulture's checklist and observation tools are particularly useful for structuring these kinds of warehouse coaching and operator observation sessions — creating consistent frameworks that team leaders can use repeatedly, and building a documented record of what good warehouse picking accuracy actually looks like in practice.

4. We Made Performance Visible

Perhaps the single highest-impact change was the simplest: making warehouse performance data visible to the team in real time.

When operators can see their own performance relative to their own historical average — not ranked against colleagues, but tracked against their personal baseline — behaviour changes.

This is not about pressure or surveillance. It is about giving people the information they need to self-manage. Most operators, when they can see their own output clearly, want to improve it. The same applies to picking accuracy: when operators understand that warehouse picking errors have a direct impact on visible metrics, the incentive to get it right increases without any additional management pressure.

A simple daily performance board showing team average LPH, shift LPH and individual trend direction (up, down, stable) was enough to create the feedback loop that sustained improvement over time. Operations that also surface picking accuracy or order accuracy data on the same board typically see the same self-correcting behaviour around error reduction.

The Numbers After 12 Months

The improvement was not linear. The first two months showed modest gains as measurement frameworks bedded in. Months three to six showed the sharpest improvement as coaching and visibility changes took effect. Months seven to twelve showed consolidation and sustained performance at the new level.

This pattern — slow start, sharp middle, stable finish — is typical of behaviour-driven operational improvement. It is different to the pattern you see from technology-driven change, which tends to show immediate step-change followed by gradual decay if adoption is not sustained.

Warehouse Management Tools That Actually Made a Difference

Being honest about technology: the warehouse productivity improvement above was primarily driven by process change and people development, not software.

That said, the right warehouse management tools made the process change faster and more sustainable. Here is what was genuinely useful:

Three Things to Do This Week

If you are running a pick operation and want to start moving your LPH numbers and reducing warehouse picking errors, here is the minimum viable starting point:

1. Pull your last 30 days of pick data by operator

   Build a simple pivot table showing average LPH per operator. Look at the spread. If it is wide, you have a behaviour and warehouse coaching opportunity. If your WMS also tracks pick errors or mispick rates, add that column — the operators with wide accuracy gaps are your first coaching priority.
2. Split your data by shift

   If you run multiple shifts, compare LPH by shift for the same period. Any consistent difference is a signal worth investigating. The answer will tell you something specific about what is working — and usually reveals whether warehouse picking errors are concentrated in one shift context rather than spread evenly.
3. Spend one hour observing your top performer

   Not to put them on a pedestal — to understand what they are doing differently. Note their pick path habits, scanner technique and how they handle confirmations. Document it. This becomes the practical foundation of your warehouse coaching framework for improving both speed and picking accuracy across the team.

None of these require budget, software or sign-off. They require a spreadsheet and time.

Warehouse Error Reduction and Picking Efficiency: The Same Problem

Improving warehouse picking accuracy and reducing picking errors is not a separate initiative from improving warehouse efficiency. It is the same problem approached from both directions.

The 21% improvement documented here came from a team that was already working hard. The gain came from working smarter — with better warehouse analytics, clearer performance feedback and more deliberate operator development. It demonstrates that you can meaningfully improve warehouse productivity, improve picking accuracy and reduce warehouse picking errors without new equipment, without a WMS upgrade and without capital approval.

Warehouse continuous improvement does not require a big project. It requires measurement, honesty about what the data is telling you, and the discipline to follow through.

That is available to any pick operation willing to look at the numbers seriously.