Everyday Reference
Terms and habits for people who read dashboards, not build them.
You shouldn't need a data science degree to sit in a meeting confidently. This page collects the vocabulary and small habits that make reports less intimidating.
Glossary
Terms you'll meet in almost any dashboard.
KPI
Key Performance Indicator. A metric explicitly tied to a business goal, tracked because a decision depends on it.
Metric
Any measurable number. Every KPI is a metric, but most metrics are not important enough to be a KPI.
Dimension
A category used to break a number down, such as region, product line or customer type.
YoY / MoM
Year-over-year and month-over-month. Shorthand for comparing a figure against the same period previously.
Trend line
A smoothed line drawn through data points to show general direction, which can obscure short-term volatility.
Outlier
A data point far from the rest, which can be a genuine event or a data entry error worth checking first.
Data source
Where a number originates, such as a booking system or a form. Two dashboards can disagree if they pull from different sources.
Refresh rate
How often a dashboard updates. A "live" dashboard refreshed once a day is not actually live in the moment you're viewing it.
Habits that build confidence over a few months
Keep a running note of any term used in a meeting that you didn't fully understand, and look it up afterwards rather than nodding along. Ask the person presenting a chart what would need to be true for the number to be wrong, not just what the number is.
When a report claims something changed "because of" an action taken, ask whether anything else happened at the same time that could explain it. Correlation gets treated as causation constantly in business reporting, usually without anyone intending to mislead.
On running a dashboard review without dominating it.
If you're the one leading a review meeting, resist opening with your own interpretation of the numbers. Let the room look first. People read a chart differently when they haven't already been told what conclusion to reach, and disagreement at this stage is useful, not a sign the meeting is going badly.
Close every review with a decision, even a small one, rather than just an update. "We'll check this again next week" counts. A dashboard reviewed without any resulting decision tends to stop getting reviewed at all within a couple of months.
Common Mistakes
Patterns worth recognising in your own reporting.
Metric creep
A dashboard that started with three KPIs slowly grows a fourth, fifth and sixth "just in case." Six months later nobody remembers why half of them are there. Revisiting the dashboard quarterly and removing anything unused keeps it useful.
Reacting to noise
Daily figures often swing for reasons that have nothing to do with performance. Treating a single bad day as a crisis, before checking whether it sits within normal variation, wastes energy that a weekly or monthly view would have saved.