Control charts for small businesses: spot real change

Key takeaways

A control chart plots a measure over time with a centre line and upper and lower limits based on its normal variation. Points inside the limits are usually noise; points outside, or long runs on one side, signal a real change. It stops businesses overreacting to one bad week and helps prove that an improvement, such as outsourcing a process, really worked.

• Plot a measure weekly with a centre line and limits.
• Inside the limits is usually normal variation.
• Outside the limits, or a long run on one side, signals change.
• Use it to prove an improvement is real.

What is a control chart?

A control chart is a line chart of a process measure over time, with a centre line at the average and control limits above and below, usually set at three standard deviations. It was developed by Walter Shewhart at Bell Laboratories in the 1920s and is the core tool of statistical process control.

• How outsourcing can improve business efficiency
• Operations dashboard: the few numbers a small firm needs

Why does it matter?

Every process varies from week to week. Without a control chart, managers react to normal ups and downs as if they were problems, which wastes time and unsettles teams. A control chart separates common-cause variation, the normal noise, from special-cause variation, a real change worth acting on.

• The cost of micromanagement in a small business
• Leading vs lagging indicators in operations

How do you build one?

Take at least 15 to 20 weeks of a measure, such as right-first-time rate or lead time. Calculate the average and the standard deviation. Draw the centre line at the average and the limits at three standard deviations above and below. Plot each new week as it arrives. A spreadsheet is enough.

• 1. Collect 15 to 20 weeks of one measure.
• 2. Calculate average and standard deviation.
• 3. Draw centre line and limits.
• 4. Plot each new week.
• 5. Look for points outside limits or long runs.

• Outsourcing efficiency KPIs: prove the work got faster
• Right first time rate: the quality number that cuts cost

How do you read it?

A point outside the limits signals something unusual that week. A run of about seven or eight points on one side of the centre line suggests the process has shifted. A steady trend up or down over several weeks suggests drift. Anything else is usually normal variation, so leave it alone.

• Throughput: how much work your team really finishes
• Continuous improvement in outsourcing: savings past year one

What does it look like in practice?

In an illustrative example, weekly lead time for invoices averages four days with limits of two and six days. After a process is handed to a managed team, the next nine weeks all sit below the centre line, between one and two days. That run confirms a real shift, not a lucky few weeks. Example figures only.

• Post-implementation review: did outsourcing work?
• How to measure the time outsourcing saves you

When is a control chart overkill?

It is overkill for measures that change rarely, for processes with very few items a week, or when a simple trend line already makes the picture clear. Use it where people argue about whether a change is real.

• Work sampling: measure where office time goes, quickly

What does this look like in practice?

A pattern we see: a manager calls a meeting every time a weekly figure dips, and the team spends time explaining noise. A control chart shows most dips are within normal range, and the meetings stop until a real signal appears.

• See client work
• BPO Solutions for UK businesses

Control chart checklist

Use these to get started.

• Pick one important weekly measure.
• Collect 15 to 20 weeks of data.
• Draw the centre line and limits.
• Act only on clear signals.
• Recalculate limits after a real change.

• Outsourcing efficiency checklist: 20 checks

Next step

Send us 15 to 20 weeks of one measure. We will build a simple control chart and show you whether recent changes are real.