🔍 FDA Food Incidents Analysis — SQL Project

Exploratory SQL analysis of 231,897 food-related adverse event reports from the U.S. Food & Drug Administration (FDA) — identifying industry risk patterns, consumer profiles, and year-over-year trends using Google BigQuery.

---

🧠 Business Context

In the FMCG industry, monitoring product safety incidents is critical for commercial and supply chain teams. This project simulates the kind of analytical work done in commercial performance roles — identifying where incidents concentrate, who is affected, and how trends evolve over time.

Dataset: bigquery-public-data.fda_food.food_events Tool: Google BigQuery (SQL) Skills demonstrated: Aggregation, filtering, multi-column grouping, date extraction, CTEs, INNER JOIN, YoY analysis

---

📊 Key Findings

Metric	Result
Total incidents analyzed	231,897
Distinct brands involved	84,086
Product categories covered	45
#1 industry by incidents	Vitamins/Minerals/Proteins (95,043)
#2 industry by incidents	Cosmetics (83,526)
Peak year for Cosmetics	2021 (21,247 incidents)
Most affected consumer profile	Females aged 53–60
Fastest growing category 2022→2023	Baby Food Products (+92 incidents, +109%)
Largest decline 2022→2023	Nuts/Edible Seeds (–1,710 incidents)

---

🗂️ Analysis Structure — 5 Queries

Query 1 — Global Overview

Question: What is the overall scale of the dataset?

SELECT
  COUNT(report_number) AS nombre_incidents,
  COUNT(DISTINCT products_brand_name) AS nombre_marques,
  COUNT(DISTINCT products_industry_name) AS nombre_categories
FROM `bigquery-public-data.fda_food.food_events`

Result: 231,897 incidents · 84,086 distinct brands · 45 product categories

---

Query 2 — Top 10 Industries by Incident Volume

Question: Which food industries concentrate the most incidents?

SELECT
  products_industry_name,
  COUNT(report_number) AS nombre_incidents
FROM `bigquery-public-data.fda_food.food_events`
GROUP BY products_industry_name
ORDER BY nombre_incidents DESC
LIMIT 10

Top 3 results:

1. Vitamins/Minerals/Proteins — 95,043 incidents
2. Cosmetics — 83,526 incidents
3. Nuts/Edible Seeds — 5,760 incidents

---

Query 3 — Temporal Evolution by Industry

Question: How do incidents evolve year by year per industry?

SELECT
  products_industry_name,
  EXTRACT(YEAR FROM date_created) AS annee,
  COUNT(report_number) AS nombre_incidents
FROM `bigquery-public-data.fda_food.food_events`
GROUP BY products_industry_name, annee
ORDER BY nombre_incidents DESC

Key insight: Cosmetics peaked in 2021 with 21,247 incidents — nearly 3x higher than any other year, suggesting a specific market event or reporting surge.

---

Query 4 — Consumer Profile Analysis

Question: Who are the most affected consumers — by gender and age?

SELECT
  consumer_gender,
  consumer_age,
  COUNT(report_number) AS nombre_incidents
FROM `bigquery-public-data.fda_food.food_events`
WHERE consumer_gender IS NOT NULL
  AND consumer_age IS NOT NULL
GROUP BY consumer_gender, consumer_age
ORDER BY nombre_incidents DESC
LIMIT 5

Key insight: The top 5 most affected profiles are all Female, aged 53–60 — suggesting strong concentration of incidents in cosmetics and dietary supplements targeting this demographic.

---

Query 5 — Year-over-Year Comparison 2022 vs 2023 (CTE + JOIN)

Question: Which industries are growing or declining in incident volume?

WITH incidents_2022 AS (
  SELECT
    products_industry_name,
    COUNT(report_number) AS incidents_2022
  FROM `bigquery-public-data.fda_food.food_events`
  WHERE EXTRACT(YEAR FROM date_created) = 2022
  GROUP BY products_industry_name
),
incidents_2023 AS (
  SELECT
    products_industry_name,
    COUNT(report_number) AS incidents_2023
  FROM `bigquery-public-data.fda_food.food_events`
  WHERE EXTRACT(YEAR FROM date_created) = 2023
  GROUP BY products_industry_name
)
SELECT
  a.products_industry_name,
  a.incidents_2022,
  b.incidents_2023,
  b.incidents_2023 - a.incidents_2022 AS evolution
FROM incidents_2022 a
INNER JOIN incidents_2023 b
  ON a.products_industry_name = b.products_industry_name
ORDER BY evolution DESC

Key insights:

• 📈 Baby Food Products saw the sharpest increase: +92 incidents (+109%) — a critical signal for FMCG players in infant nutrition
• 📉 Nuts/Edible Seeds saw the largest decline: –1,710 incidents, suggesting improved safety controls or market contraction
• 📉 Cereals/Breakfast Food dropped by –617 incidents — relevant for brands like Danone operating in this category

---

⚙️ Technical Skills Demonstrated

Skill	Used in
SELECT, FROM, WHERE	All queries
COUNT(), COUNT(DISTINCT)	Queries 1, 2, 3, 4, 5
GROUP BY (single & multi-column)	Queries 2, 3, 4, 5
ORDER BY, LIMIT	Queries 2, 3, 4
EXTRACT(YEAR FROM date)	Queries 3, 5
IS NOT NULL filtering	Query 4
CTE (WITH ... AS)	Query 5
INNER JOIN	Query 5
YoY delta calculation	Query 5

---

🔗 Related Projects

• Power BI Commercial Dashboard — Multi-page interactive BI report · €120M scope · Danone SSA
• Excel Commercial BI Dashboard — 8-tab Excel dashboard · 8 brands · 70+ distributors
• VBA ETL Multi-File Compiler — Automated ETL processing 70+ files in under 4 minutes

---

👤 Author

Serge NGUENA — Senior Commercial Performance & BI Analyst LinkedIn · GitHub Portfolio · Laval, QC