Procter and Gamble India Data Consultant Interview: Rounds, Questions and Preparation

What the job involves

A Data Consultant in GCCs, consulting, and the Big 4 helps clients turn raw data into business decisions by cleaning, analysing, and visualising data. Day to day work includes writing SQL queries, building dashboards, documenting findings, and presenting recommendations to client teams. Freshers often support senior consultants on data audits, KPI design, and process automation projects.

Role and technical knowledge

1. 1

   How would you explain the difference between a primary key and a foreign key in a client database?

   Define each clearly, give a small example like customer ID versus order ID, and mention why it matters for joins and data integrity.

2. 2

   What steps do you follow to clean a messy sales dataset before analysis?

   Cover removing duplicates, handling missing values, standardising formats, checking outliers, and documenting every change for auditability.

3. 3

   When would you choose a bar chart over a line chart in a client dashboard?

   Explain that bar charts compare categories while line charts show trends over time, and mention matching the chart to the client's decision.

4. 4

   What is the difference between a KPI and a metric in a consulting context?

   State that a KPI is a metric tied directly to a strategic goal, while a metric is any measurable value, and give a retail example like revenue per store versus total footfall.

Practical scenarios

1. 5

   A client says their monthly report takes three days to prepare manually. How would you approach automating it?

   Start by mapping the current steps, identify repetitive tasks, suggest Excel macros or a Python script, and propose a validation step before sharing results.

2. 6

   You find that two client teams report different numbers for the same sales figure. What do you do?

   Trace the data sources, check definitions and filters, document the discrepancy, and recommend a single source of truth with clear definitions.

3. 7

   Your manager asks you to build a dashboard for a client who has never used one. How do you ensure adoption?

   Focus on simple visuals, clear labels, a short training session, and gathering feedback after the first week to iterate.

4. 8

   You are given a dataset with 30 percent missing values in a key column. What are your options?

   Discuss dropping rows only if safe, imputing with mean or median, using a separate category for missing, and flagging the limitation in your report.

Behavioural and HR

1. 9

   Tell me about a time you had to explain a technical finding to someone non technical.

   Use a project example, describe how you simplified the language or used a visual, and end with the outcome or decision it enabled.

2. 10

   Why do you want to work as a data consultant in a GCC or Big 4 environment?

   Connect your interest in data with exposure to multiple clients, structured problem solving, and the learning curve of consulting projects.

3. 11

   Describe a situation where you received critical feedback on your work.

   Show that you listened, asked clarifying questions, made a specific change, and improved the final output.

4. 12

   How would you handle a client deadline that clashes with a team commitment?

   Emphasise early communication, prioritising based on impact, offering a trade off, and keeping both parties informed without overpromising.

Consulting and sector awareness

1. 13

   What does a data consultant actually deliver to a client at the end of a project?

   Mention cleaned data, dashboards, documented insights, a slide deck with recommendations, and a handover session for the client team.

2. 14

   How is a GCC data role different from a core analytics role in a product company?

   Highlight the client facing nature, variety of projects, emphasis on documentation and storytelling, and exposure to multiple business domains.

3. 15

   What are common data quality issues you expect in Indian retail or FMCG client data?

   Talk about inconsistent store codes, duplicate entries, missing dates, mixed units, and manual entry errors, then suggest validation checks.

4. 16

   If a client asks for a prediction but you only have three months of data, what do you recommend?

   Explain the limitation, suggest descriptive analysis and simple trends instead, propose collecting more data, and set clear expectations.