Understanding Your Customers

Disclaimer: Super Furniture is a fictional company used for this example. Below is a showcase of how HyperData can be used. Data included in this example are simulated and for demonstration purposes.

Problem

Super Furniture is an e-commerce store that sells furniture and home goods. The company is interested in better understanding its customers to drive top-line growth. To do so, Super Furniture is looking to leverage 3rd party data to enable granular customer segmentation, targeted advertising campaigns, personalized messaging, and market research.

The problem is that it is incredibly tedious & expensive to extract, transform, and validate large volumes of 3rd party data from different sources. It requires weeks to months of data engineering work and ongoing data pipeline maintenance to get real value from the data - this is an incredibly high upfront cost.

Solution

Super Furniture can use the HyperData API to simplify the data integration process and significantly reduce costs. HyperData transforms and validates the raw 3rd party data so that clean data can be loaded into Super Furniture's database in a unified and reliable format. HyperData does this at a fraction of the cost and time vs. Super Furniture building and maintaining the integrations in-house.

Source Data

Note that the data is simulated and is for demonstration purposes only.

Orders

The order data comes directly from the Super Furniture database and contains the records of each online order. We will use the zipcodes and emails from the transaction records to learn more about the customer base.

106KB

orders.csv

Here's a sample row from the dataset:

customer_name	customer_address	purchased_at	total_amount	customer_email	zipcode	city	state	device_type	order_id
Lindsay Lee	6881 Encolumn Blvd	2024-03-29	213.61	[e-mail]	66984	Fort Worth	TX	tablet	order_1000

Zip Code

To better understand their customer base, Super Furniture has purchased a zip code dataset detailing some key statistics on the local population in each zip code. This dataset will be used to better understand where the customers are coming from.

162KB

zipcode.csv

Here's a sample row from the dataset:

zipcode	average_household_income	population	average_people_in_household	average_age	most_frequent_dwelling_type	average_credit_score	consumer_spending_score	median_home_value	average_rent_per_sq_ft	median_commute_time	most_common_education_level	average_number_of_vehicles_per_household	average_number_of_pets_per_household	collection_date
25617	67794.12	18263	2.8	45.5	apartment	740.9	13	329501.76	2.43	26	bachelors	1.42	0.49	2023-01-01

Email

To dive deeper into understanding each customer individually, Super Furniture has purchased an email enrichment dataset from a 3rd party that has appended a bunch of key attributes to each email. This dataset will be used to better understand each customer.

77KB

email.csv

Here's a sample row from the dataset:

email	age_group	gender	intrest_tags	engagement_score	open_rate	social_media_activity	frequently_visited_sites	date_collected
[e-mail]	55-64	male	fashion	83.8	40.4	low	entertainment	2021-02-10

Transforming with Hyperdata

The source data is in different formats and will be updated at different frequencies so we will use the HyperData API to transform the source data into a single format for use. Here's the code we will use to do this:

import hdata as hd

# Load the raw data
orders = hd.Source(transformation_key='*****', data='orders.parquet')
zipcode = hd.Source(transformation_key='*****', data='zipcode.csv')
email_enrichment = hd.Source(transformation_key='*****', data='email.csv')

# Transform
sources = [orders, zipcode, email_enrichment]
entity, attribute, record = hd.transform(sources, auth_token='*****')

Here is the output of the transformation in CSV format for reference. The data is returned from the transform function as three pandas data frames.

entity

156KB

entity.csv

attribute

2KB

attribute.csv

record

3MB

record.csv

With this clean data feed, Super Furniture can now run targeted advertising campaigns, customize communications with existing customers, and better understand their market. All of this is made possible by creating a single source of truth for external data with the HyperData API.