Business Metrics Deep-Dive Report
Part of the Data Analyst Roadmap
In short
"Revenue dropped" is where most people stop. Breaking it into the parts that make it up, and finding which part moved, is what turns an analyst into the person leaders call first.
Pick one business metric (e.g. churn, conversion) and analyze it thoroughly — trend over time, key drivers, and segment breakdowns — showing depth on a single question rather than surface coverage of many.
Think of it like this
The scooter mechanic
Your scooter feels slow. A good mechanic doesn't just say "it's slow". He checks fuel, spark plug, air filter and tyre pressure one by one, until he finds the part that's off. A metrics deep dive checks each part of a number the same way.
Watch it happen
Breaking revenue into its parts
1/5Revenue fell 10% month on month. On its own, that tells you nothing about why.
The big ideas
What you need to know
Decompose the metric
Write the metric as a product or sum of parts. The part that moved is where you dig.
Then segment
Split the moving part by city, channel, device or customer type to find where it's concentrated.
Compare fairly
Use the same number of days and similar seasons. A short month or a festival can explain a lot.
Write it as a report
One page: what changed, which part drove it, likely cause, and what to do. Charts support each step.
See it in code
Walk through the code
1SELECT2 DATE_TRUNC('month', order_date) AS month,3 COUNT(DISTINCT customer_id) AS customers,4 COUNT(*) * 1.0 / COUNT(DISTINCT customer_id) AS orders_per_customer,5 SUM(amount) / COUNT(*) AS avg_order_value,6 SUM(amount) AS revenue7FROM orders8WHERE status = 'delivered'9GROUP BY 110ORDER BY 1;1/3Unique customers who ordered that month.
One query gives all three parts of revenue by month. Multiply them and you get revenue back, which is a handy check.
Save yourself the pain
Common mistakes
Oops
Jumping to a cause like "the new competitor" before decomposing the metric.
Do this instead
Break the number into parts first. The data often points somewhere you didn't expect.
Oops
Comparing October with a 31-day month against a short or festival-heavy month without saying so.
Do this instead
Use daily averages or same-period comparisons, and call out calendar effects.
Your mission
First milestone: decompose one month
Quick check
Revenue is flat, but customers rose 20%. Order value is unchanged. What must have happened?
Go deeper
- 1Choose a metric with enough public or sample data to genuinely dig into.
- 2Break it down by at least 2-3 dimensions (time, segment, channel).
- 3End with a specific, actionable recommendation, not just a description of the trend.
More projects at this stop
End-to-End SQL + BI Dashboard Project
Exploratory Data Analysis (EDA) on a Public Dataset
A/B Test Analysis Case Study
Data Cleaning Case Study (Messy Dataset)
Python + Pandas Analysis Notebook
Portfolio Website or GitHub Showcase
Stuck on this topic? Ask an Insider
Get 1:1 guidance from people who've walked this exact path — on the InsideEdge app.
