What Product Sense Interview Questions at Google and Meta Actually Mean for Data Analysts

For years, data analyst interviews were predictable — write some SQL, solve a probability problem, maybe walk through a case study. But Google and Meta have significantly evolved their interview formats over the last two years. Product sense questions have become a core part of the data analyst interview loop at both companies, and this shift is now trickling down to Indian tech giants like Flipkart, Swiggy, Paytm, and Meesho as well.

So what exactly is a “product sense” question for a data analyst? Unlike a PM interview where you’re asked to design a product from scratch, a data analyst’s product sense question typically asks you to evaluate a product decision using data. Think questions like: “Daily active users on Google Search dropped 10% last week — how do you investigate this?” or “Meta is testing a new Reels recommendation algorithm — what metrics would you track to evaluate its success?”

The underlying skill being tested is your ability to connect business context to data decisions. Interviewers at Google and Meta want to see that you understand why a metric matters, not just how to calculate it. They want you to think about user behaviour, business goals, and the potential trade-offs in any product decision. This is fundamentally different from just being a SQL wizard, and it’s the gap that trips up even the most technically strong candidates.

In India, this trend is accelerating fast. Companies like Zomato, PhonePe, and CRED have started borrowing heavily from Google and Meta’s interview playbooks. If you’re preparing for a senior data analyst or lead analyst role anywhere in the Indian tech ecosystem in 2026, product sense is non-negotiable. The good news? It’s a learnable skill — and this post will show you exactly how to build it.

Real Product Sense Interview Questions You’ll Face at Google, Meta, and Top Indian Tech Companies

These aren’t hypothetical — these are the types of questions that have actually been asked in data analyst loops at Google, Meta, and their Indian counterparts like Flipkart and Swiggy. Practising these will build the muscle memory you need to think on your feet. Notice how each question blends product thinking with data analysis — that’s the key pattern to internalise before your interview.

1. “Google Maps noticed a 15% drop in navigation sessions in Tier-2 Indian cities over the last 30 days. How would you diagnose the root cause, and what data would you pull first?”
2. “Meta is considering removing the ‘like count’ from public posts in India to reduce social comparison anxiety. How would you design an A/B test to measure the impact, and what guardrail metrics would you set to protect core engagement?”
3. “Swiggy’s average order value has increased by 8% month-over-month, but customer retention has dropped 5%. Is this a good or bad sign for the business, and how would you investigate whether these two trends are related?”
4. “You are a data analyst at Google Pay India. A new UPI cashback feature launched two weeks ago shows strong click-through rates but weak transaction completion. Walk me through your analysis plan.”
5. “Meta’s Feed team wants to prioritise ‘meaningful social interactions’ over passive consumption. How would you operationally define ‘meaningful’ as a data analyst, and what metrics would you propose to track progress toward this goal?”

The SQL Skill That Powers Product Sense Answers: Funnel Analysis

One of the most common underlying analyses in product sense interviews is funnel analysis — understanding where users drop off in a product flow. Whether it’s a Google Search journey, a Meta Reels watch session, or a Paytm payment flow, being able to write clean, efficient funnel analysis SQL is essential. Here’s a practical example: analysing a conversion funnel for a feature like Google Pay’s UPI cashback flow, where you want to understand step-by-step drop-off rates from impression to transaction completion.

-- Funnel Analysis: UPI Cashback Feature Drop-off
-- Table: user_events (user_id, event_name, event_timestamp, session_id)

WITH funnel_steps AS (
  SELECT
    user_id,
    -- Step 1: User saw the cashback banner
    MAX(CASE WHEN event_name = 'cashback_banner_impression' THEN 1 ELSE 0 END) AS step1_impression,
    -- Step 2: User clicked on the cashback offer
    MAX(CASE WHEN event_name = 'cashback_offer_click' THEN 1 ELSE 0 END) AS step2_click,
    -- Step 3: User initiated a UPI transaction
    MAX(CASE WHEN event_name = 'upi_transaction_initiated' THEN 1 ELSE 0 END) AS step3_initiated,
    -- Step 4: User completed the UPI transaction
    MAX(CASE WHEN event_name = 'upi_transaction_completed' THEN 1 ELSE 0 END) AS step4_completed
  FROM user_events
  WHERE event_timestamp >= DATE_SUB(CURRENT_DATE(), INTERVAL 14 DAY)
  GROUP BY user_id
),

funnel_counts AS (
  SELECT
    SUM(step1_impression)  AS total_impressions,
    SUM(step2_click)       AS total_clicks,
    SUM(step3_initiated)   AS total_initiated,
    SUM(step4_completed)   AS total_completed
  FROM funnel_steps
)

SELECT
  total_impressions,
  total_clicks,
  total_initiated,
  total_completed,
  -- Drop-off rates at each step
  ROUND(100.0 * total_clicks / NULLIF(total_impressions, 0), 2)   AS click_rate_pct,
  ROUND(100.0 * total_initiated / NULLIF(total_clicks, 0), 2)     AS initiation_rate_pct,
  ROUND(100.0 * total_completed / NULLIF(total_initiated, 0), 2)  AS completion_rate_pct,
  -- Overall end-to-end conversion
  ROUND(100.0 * total_completed / NULLIF(total_impressions, 0), 2) AS overall_conversion_pct
FROM funnel_counts;

In a product sense interview, this kind of query does double duty — it demonstrates technical SQL fluency AND shows the interviewer that you think in terms of user journeys, not just individual data points. When you present this output in an interview, walk through what each drop-off rate means for the product decision at hand. That’s what separates a good data analyst answer from a great one.