Control charts
A control chart shows results in time order with limits worked out from the process itself, so you can tell noise from a signal.
About 5 minutes to read.
Key idea
A control chart plots results in the order they happened, with a centre line at the average and two limits, usually three standard deviations either side, calculated from the data. The limits are the voice of the process: they describe what the process does on its own, not what anyone wishes it did.
Points scattered between the limits, with no long runs or trends, are ordinary variation. Leave them alone. A point outside the limits, a long run on one side of the centre, or a steady climb is a signal that something has changed. That is the moment to look for the cause, while the trail is still warm.
The limits are not targets. A target belongs to the customer or the business. Mixing the two is one of the most common mistakes in reporting.
Example: call answering
The Tuesday spike
A council contact centre charts the average time to answer each day. For weeks the points wander between 40 and 90 seconds, inside limits of 20 and 110. Then one Tuesday shows 160 seconds.
That point is a signal. A quick look finds a new council tax letter that went out on Monday with the wrong phone number for a different service. Fixing the letter fixes the spike. Had the team reacted to every ordinary day above the average, they would have been chasing ghosts.