Challenge #1: Ecommerce Insights

Background

With the holidays approaching, your online business needs to be more data driven than ever!

Amazon Package GIF

You have access to ecommerce orders from 2023-2024 in the attached dataset (source: Kaggle). You’re looking for insights on customer satisfaction over time for electronics goods sold to inform your future strategy.

ecommerce_customer_behavior_dataset.csv (550.0 KB)

Data Challenge

For first-time customers purchasing electronics goods, how did satisfaction vary by month? Respond in the forum below with a screenshot of your pipeline to complete the challenge!

Here’s the visualization you need to create:

Tips & Tricks

First time uploading data to Prophecy? Follow these steps:

How to upload a dataset

(using the Starter/Professional edition)

  1. On the Prophecy home page shortcuts, select “New Project”

  2. Name your project, select your team and “prophecy for analysts” template. Select the default fabric

  3. Create and name a new pipeline “ecommerce_1”

  4. Drag and drop the attached .csv file onto the canvas

  5. Select the default database, set the schema to “main,” and type “ecommerce” into the table dropdown to create a new table

  6. Click “next” and then “done” to finish uploading the file

Are you stuck? Open the hint below:

Hint

Make sure you filter on returning customers and electronic goods only. Then you’ll need to create the missing month column and use an aggregate gem!

Solution

When you’ve finished the challenge, here’s the expected data explorer result:

Solution

Visualization configuration:

Share a screenshot of your completed pipeline below to earn a challenge badge!

What a fun challenge @MeganBowers! It took me a couple of iterations to manually build it (I forgot to filter on Electronics!) :melting_face:

I then enjoyed watching the Prophecy AI Copilot build the pipeline from my prompt. :exploding_head:

Looking forward to future challenges!

1 Like

@Callum_Finlayson great work, and glad you enjoyed it! Love that you were able to build the pipeline from a prompt as well!