01

Name the metric and the period

“How are we doing?” can hide several different questions. Specify revenue, margin, active customers or another metric, along with the time period and the comparison you care about.

02

Separate a change from its explanation

A chart can show that sales increased. It cannot, by itself, tell you why. Start by describing the change, then investigate price, volume, customer mix, seasonality and other plausible contributors.

03

Look beneath the total

A healthy headline can conceal a weak segment. Compare regions or products, check whether definitions are consistent and account for missing or delayed records before interpreting a difference.

04

Keep a route back to the data

A narrative summary is useful when the underlying values can be inspected. Keep the time range, filters, units and data freshness visible next to the answer.

05

End with a next question

An analysis should narrow the uncertainty around a decision. Write down what is known, what remains uncertain and the next piece of evidence that would be most useful.

Put a question into practice.

Explore a sample workflow in the Cynaris demo centre.

Try a product demo