Reporting

Packaging the Analysis for People Who Will Act on It

Reporting is where analysis becomes communication. The numbers, charts, and comparisons from the previous steps are assembled into a structured artifact — a dashboard, a slide deck, or a written report — designed for a specific audience and a specific decision. A good report is not a data dump; it is concise, leads with the metrics that matter, and ends with a recommendation someone can act on.

Reporting consolidates the analysis into one artifact built for a specific audience and decision.

Know the Audience First

• Executives want the headline, the trend, and the recommendation — one screen, no clutter.
• Operational teams want detail they can act on — by region, by SKU, refreshed often.
• Analysts want the methodology and the ability to drill down into the underlying data.

Anatomy of a Good Report

1. Headline metrics (KPIs) at the top — the few numbers that define success.
2. Trend and comparison visuals that give those numbers context.
3. A short narrative: what happened, why, and what to do about it.
4. Supporting detail for those who need to dig deeper, kept out of the main flow.

In Python — a summary table for the report

Python · pandas

import pandas as pd

# One consolidated summary table by region and category
summary = (orders.groupby(["region", "category"])
                 .agg(sales=("sales", "sum"),
                      profit=("profit", "sum"),
                      orders=("order_id", "nunique"))
                 .assign(margin=lambda d: (d["profit"] / d["sales"]).round(3))
                 .sort_values("sales", ascending=False)
                 .reset_index())

# Headline KPIs for the top of the report
kpis = {
    "total_sales": orders["sales"].sum(),
    "total_profit": orders["profit"].sum(),
    "overall_margin": orders["profit"].sum() / orders["sales"].sum(),
}

# Export the table for a dashboard or document
summary.to_html("report_summary.html", index=False)
print(kpis)

In SQL — a roll-up with subtotals

SQL

-- GROUPING SETS produces detail rows AND subtotals in one pass,
-- exactly what a report's summary table needs.
SELECT
  region,
  category,
  SUM(sales)  AS sales,
  SUM(profit) AS profit
FROM orders
GROUP BY GROUPING SETS (
  (region, category),   -- detail
  (region),             -- per-region subtotal
  ()                    -- grand total
)
ORDER BY region, category;

Tools Commonly Used

• Python: Pandas groupby.agg, Styler, to_html/to_excel; Jupyter notebooks.
• SQL: GROUP BY GROUPING SETS, ROLLUP, CUBE for subtotals.
• BI: Power BI and Tableau dashboards; presentation tools for executive decks.

Best Practices

• Lead with the answer, not the methodology — put the recommendation where it will be read.
• Keep one idea per chart; resist the urge to crowd a dashboard.
• State assumptions, the date range, and data freshness so the report is trustworthy.
• Automate recurring reports so the numbers are always current and never hand-copied.

Example: The Retail Performance Report

For the retail dataset, the report might consist of:

• Three KPI cards: total sales, total profit, and overall margin, each with a YoY delta.
• A sales-by-region-and-category table with subtotals.
• A short narrative flagging the one region that is growing revenue while losing margin.