Descriptive Analytics
Every Monday Nisha wants one thing before her coffee: "How did last week go?" That question does not ask why, and it does not ask what comes next. It asks for an accurate summary of the past, and it is the most common request an analyst gets. This section follows one week of cold-coffee sales:
| day | Mon | Tue | Wed | Thu | Fri | Sat | Sun |
|---|---|---|---|---|---|---|---|
| cups | 40 | 38 | 36 | 43 | 46 | 95 | 38 |
The rule
Descriptive analytics answers "What happened?", nothing more.
It turns many rows into a few facts. Its tools are simple:
- Totals and counts: 336 cups sold in the week.
- Averages: 336 ÷ 7 = 48 cups a day.
- Extremes: the best day was Saturday (95), the quietest Wednesday (36).
- Comparisons: this week against last week, or one item against another.
- Shares: Saturday alone was 95 ÷ 336, about 28% of the week.
Dashboards, weekly reports and KPI tiles are all descriptive analytics. They are the base the other three types build on: you cannot explain, forecast or recommend without first knowing what happened.
The shortcut
Check the verb tense of the question. Past tense and "how many / how much" means descriptive. Then give three things: the headline number, a comparison, and the standout.
Headline number
Total: 40 + 38 + 36 + 43 + 46 + 95 + 38 = 336 cups.
The trap is sneaking a cause or a prediction into a descriptive report. "Sales rose because of our new menu" is a claim you have not tested.
Drills
Answer fast — trust the rule.
What was the total number of cups sold in the week?
Which question is descriptive?
Which statement goes beyond descriptive analytics?
Last week sold 300 cups and this week 336. What is the percentage change?
Remember
- Descriptive analytics answers "What happened?", nothing more.
- Shortcut: past tense means descriptive; give the headline, a comparison and the standout.
- Trap: sneaking in a "because" or a "next week" you have not tested.