I recently got this question in a Snowflake interview and wanted to share it.

Problem Statement

You are given two datasets:

• Products (as a vector of pairs):
  Each entry contains:
  product_id
  price

• Sales (as a list of records):
  Each entry contains:
  customer_id
  product_id
  quantity
  Task

For each product, compute:

Number of unique customers who bought the product and also if one customer bought multiple product or if one prodcut was bought 1 guy and same product was bought by another guy then it should say 2 guy bought this product with the total revenue generated by the prodcut

Return/output in the format:

<number_of_customers> <product_id> <total_revenue>

• Approach

Convert the products vector into a hashmap:

product_id → price
Iterate over sales:
Lookup price using product_id
Compute revenue = price × quantity
Aggregate revenue per product
Track unique customers using a set
Output results per product

// Given as vector of pairs
vector<pair<int, int>> products = {
{101, 10},
{102, 20},
{103, 15}
};

// sales: {customer_id, product_id, quantity}
vector<tuple<int, int, int>> sales = {
{1, 101, 5},
{2, 103, 2},
{1, 103, 1}
}

// Convert to map
unordered_map<int, int> priceMap;
for (auto &p : products) {
    priceMap[p.first] = p.second;
}



unordered_map<int, long long> revenueMap;
unordered_map<int, unordered_set<int>> customersMap;

// Process sales
for (auto &[customer_id, product_id, quantity] : sales) {
    int price = priceMap[product_id];
    long long revenue = price * quantity;

    revenueMap[product_id] += revenue;
    customersMap[product_id].insert(customer_id);
}

// Output
for (auto &[product_id, revenue] : revenueMap) {
    cout << customersMap[product_id].size() << " "
         << product_id << " "
         << revenue << endl;
}

return 0;

}